Professional Documents
Culture Documents
Scientific Knowledge and Its Social Problems
Scientific Knowledge and Its Social Problems
-Iz
(,/)0 ::AlrJ
{./lIz
go r-r-1
...... ,
f=~IZ "r ':.•
:.Y _" . :.:'V': _ .'. :.:'V': _•
o r l....,
-:"Il .....
\.~
,,:....·,:.~"r ~ "f ':.~ 'j
.~
:,
\ •... ..~/.:.\..~
I..... -...J ~,
,""
" ,,~...,
"" ":, ."" ":, ."" ,.
~.... ,,~.'" '\.i
.... ~ I · ow I....· ow I....·'
.".~~.l.
.".
,. ..•. . ' . ..... ... ' ....".. ~"l
.,. ~"'- ". ' ....,... ~"1
...., ' ....,... ".
'l\ r. :. . r.
.'.. :::0
..,.,
•• • ., • "1
'."',
,,~.'" "'"
,,~.,
••' ,r••,,\o\:\•.•• "~~\."" "~~\.""
t ...·"V I....· ow I....· ow I..
••, -...J I....· -...J "
3 0 lrJ
••, -...J i...., i....·....:.
-...J •••• "V
.. .. ~~ "~'l~'l~'
~~~\. •..,:.,.,.
~~
-.-.. • r_
:.1-·.. • :...... .o..
~_
. .""
':~_"'r ':,~."'f. ':.~j
. :.,. r.
:\.•..•..
, "1. ~'l
.·~·r···
'. " ' ,
:.,.
"~"\"" "~A
I....· -...J i....·
..,. ". :.,".- ~'l
• '."',
-..J
"
I....· ....
". ". :',.... 'R'
..,. ":..
.r•• " ' \ • '. u:
{./l r-r-1
~
":,
~
•
Transaction Publishers
New Brunswick (U.S.A.) and London (U.K.)
New material this edition copyright C 1996 by Transaction Publishers,
New Bnmswick, New Jersey 08903. Originally published in 1971 by
Oxford University Press.
All rights reserved under International and Pan-American Copyright Con-
ventions. No part of this book may be reproduced or transmitted in any
form or by any means, electronic or mechanical, including photocopy,
recording, or any information storage and retrieval system, without prior
permission in writing from the publisher. All inquiries should be ad-
dressed to Transaction Publishers, Rutgers-The State University, New
Bnmswick, New Jersey 08903.
This book is printed on acid-free paper that meets the American Na-
tional Standard for Permanence of Paper for Printed Library Materials.
Library of Congress Catalog Number: 95-32177
ISB~: I-S~8S1-2
Printed in the United States of America
Library of Congress Cataloging-in-Publication Data
Ravetz, Jerome R.
Scientific knowledge and its social problems I Jerome R. Ravetz ;
with a new introduction by the author.
p. CID.
Originally published: Oxford : Clarendon Press, 1971.9
Includes bibliographical references and index.
ISB~ I-S~851-2 (pbk. : alk. paper)
1. Science-Philosophy. 2. Science-Social aspects. I. Title.
QI75.R3 1995
SOI-dc20 95-32177
CIP
To the memory of my father
GUS RAVETZ
who taught me to think for myself
CONTENTS
THE thoughts that are developed in this book have occupied me for
such a large part of my life that to describe their origins, or even to
give credit to all those who have influenced my progress, would re-
quire an autobiographical volume. My concern for the problems of
society and ethics has been with me since childhood; and my
curiosity about the nature ofscientific knowledge was aroused at high
school. On the former problem, I at least had Lincoln Steffens'
Autobiography as an illuminating introduction; but on science, I
found the accepted 'philosophy of science' to be only another
university subject, interesting in its own right but irrelevant to the
problems that were germinating in my own mind. For some years I
had hoped that Marxism would provide a coherent synthesis of all
these problems; but in the versions then extant it was very much less
a successful docttine than an extremely general guide to action. I
recall that I was aware ofa philosophical problem-situation to which
I wished eventually to devote myself some twenty years ago when
I first came to England from America; about seven years ago I had
an opportunity to make a first statement of the problems; and since
then my work on this material has become steadily more intense, up
to the present. Through all these years, I have been mulling over
problems in my mind; and they would gain evidence from experience
in a variety of ways, mostly unsystematic and informal. My reading
and teaching in the history of science was fundamental for the
growth of my ideas; indeed, had I not been able to become an
historian rather than a mathematician, it is very unlikely that this
book could have been written. Of equal importance, however, were
the insights and instances derived from a great multitude of dis-
cussions, with friends inside and outside the academic world. Many
of these discussions were on such ordinary and practical questions,
that neither I nor my friend was aware at the time that something
\vould be 'going into my book'; and when in retrospect I would
realize that a particular idea had been suggested or stimulated by a
discussion, it was often difficult to remember which one, with which
xxvi Preface
friend, had been the occasion for the crystallizing of a vague notion.
As I have written up the text, I have tried to give credit to friends
for points which struck me at the time of discussion; I fear that this
biases the credits towards recent years, and also omits a goodly num-
ber of friends from whom I have gained a generalized wisdom.
It is possible for me to identify those institutions and organiza-
tions to which I am indebted, in various ways, for the growth of my
work. When I attended Central High School, Philadelphia, there
still remained enough of the atmORphere, and the teaching staff, of
the Gymnasium that it had once been, for me to experience the
challenge and discipline of serious academic study. Through the
generosity of the Pepsi-Cola Company in awarding me a scholar-
ship, I was able to attend Swarthmore College, where windows
opened for me. There I discovered the delights of intellectual ex-
ploration; and also found a community where people are honourable
and trustworthy in the absence of any propaganda or sanctions,
simply because of an abnosphere created by the College's Q!Iaker
traditions. Coming to Trinity College, Cambridge on a Fulbright
Scholarship, I had the privilege of apprenticeship to a master-
craftsman of mathematics, Professor A. S. Besicovitch, F.R.S. At
the (then) Durham Colleges in the University of Durham. I found
myself the protege and friend of the Warden, Sir James Fitzjames
Duff; and through him learned more of the subtle differences be-
tween English and American society. There too I met Mr. Peter
Doris, who taught' me the importance of finding the fundamental
questions and sticking to them.
When I moved to Leeds University to work with Professor
Stephen Toulmin on a Leverhulme Fellowship, I could concentrate
my energies on the history ofscience and the philosophical problems
suggested by it. The occasion for the first formal presentation of the
ideas of this book was provided by the Workers' Educational
Association; during 1962-3 I held a short course on 'science and
society', whose participants included colleagues and students from
my department and from the Department of Social Studies, and
secondary school pupils with whom I had worked in the Campaign
for Nuclear Disarmament. Just a half-century earlier, W.E.A.
teaching in Leeds had been the occasion for the achievement of a
most influential book in the philosophy of science, Norman Camp-
bell's What ;s Science? If this book is successful, it will serve as
another example of useful ideas being developed outside the context
Pre/fIG' xxvii
of teaching to a captive audience preparing for degree examinations.
For several years, a small but vigorous discussion group held under
the auspices of the Leeds University Union Student Christian
Movement examined drafts of the basic chapters of this book, and
subjected them to a sympathetically ruthless criticism. Another
small but vigorous group, a seminar in 'Social Problems of Con-
temporary Science' at the Harvard University Summer School in
1967, helped me formulate the problems discussed in the latter part
of the book. Equally important for this work has been my contact
with the peace movement in Britain, and more recently, with the
British and American societies for social responsibility in science.
Even with all these opportunities and stimuli, I would not have been
able to create this book, and give it time to mature, were I not work-
ing in the happy and encouraging environment of the University of
Leeds. My debt to the University, through successive Heads of my
department and Vice-Chancellors, together with the administrative,
clerical and teaching staff with whom I have worked, is enprmous.
The motto of the University is Bacon's 'et augebitur scientia'; I
hope that my own work' will both explain its deeper meaning, and
contribute to the fulfilment of its prophecy.
Among the individuals to whom my obligations are particularly
strong, I recall from earlier years the late ProfessorA. S. Besicovitch,
Professor C. Truesdell, Dr. D. T. Whiteside and Dr. (now Professor)
I. Lakatos, for insights into the nature of the work and of the know-
ledge achieved in the exact sciences. When Dr. D. S. L. Cardwell
and I were colleagues at Leeds, we had many conversations whose
fruits are inadequately acknowledged in the occasional citations in
the text. On the basis of those discussions, I was able to appreciate
the insights contained in Michael Polanyi's Personal Kno1lJledge,
particularly in connection with the nature and role of craft skills in
scientific work. More recently, conversations with Dr. P. M. Rattansi
provided me with the first insights on the 'craftsman's romantic
philosophy of nature' and on the category of folk-science; and from
]. E. McGuire and C. Webster I learned much concerning the
philosophical and social aspects of science during its formative
period in the seventeenth century. To R. G. A. Dolby I am indebted
for his truly. incisive criticisms of crucial sections of the text. Dis-
cussions with Dr. N. D. Ellis and Dr. A. Coddington have been
essential for the development of my ideas on· 'immature sciences';
and N. Parry exposed the obscurities at the foundations of earlier
xxviii Pre/fIG'
drafts of my argument on ethics in science. Dr. Alison Ravetz has
contributed to the work oyer the years, through sharing her own ex-
perience as an archaeologist and historian, and also providing an
asttingent and penettating stylistic criticism at several crucial points
in the evolution of the text. It was through reading the work of
Professor Barry Commoner and his group at St. Louis, that I came
to the conception of 'critical science', which gives this whole work
what unity it has. I must also thank the two typists who produced a
fm text with such speed and enthusiasm, Mrs. Lily Minkin and
Miss Margaret Knapp; and also thank the proprietors of Beck Hall,
Malham, Yorkshire, whose friendly hospitality enabled me to
achieve an intense concentration on the most difficult problems of
shaping this book, in ideal conditions. I am grateful finally to the
Oarendon ·Press for their enthusiasm for this difficult book; and
to them and the printers for producing it so quickly and well. I
hardly need add that the many imperfections of this book, some
ofthem known to me but doubdess more unknown as yet, are entirely
my own responsibility.
July 197 1 J.R.R.
INTRODUCTION
THE problem-situation that gave rise to the present work is the recent
rapid change in the character of the social activity of disciplined
inquiry into the natural world, and the consequent changes in its
understanding ofitself. The established traditions of research in 'the
philosophy of science' and in 'the sociology of science' have been
recognized as losing contact with the actual practice and real pro-
blems of science in the present period. Q!tite independendy of these
academic studies, there has developed a new common-sense under-
standing of science, derived from the daily experience of working
scientists and from the problems of decision and government within
science. I There has been a rapidly increasing flow of studies of one
or another aspect ofscience, conditioned by this new common-sense,
but none as yet that attempts a new synthesis.
It appeared to me that the problems of the character of scientific
knowledge, ofthe sociology and ethics ofscience, and ofthe applica-
tions ofscience to technology and to human welfare, are so intimately
connected that a proper study of anyone of them requires an in-
formed awareness of the others. A mixture of inherited assumptions
from the academic studies, with elements of the new common-sense;
varying in its contents from one person to the next, is not sufficient
for this. Hence I have tried to create a coherent framework of
I The most extensive codification of this new common ieDse of science is in W. O.
Hagstrom, Tile Stimtifie C"""""';'y (Basic Books, New York and London, 1965). Many
of the points made in this book can be found there, with • wealth of illustrative material
from interviews and historical examples. But his problem was very dif£el'ent from mine,
and the organization of the argument, and the emphasis given to the various points, is
quite difFerent. I felt that it would be otiose to provide a concordance of material between
the two texts, and so I am mentioning it at this early stage, with. reamunendation for its
study as a parallel source.
Rather closer to my own concerns are C. Wright Mills, Tile SotiolotittllI",.,;tllllitnl
(Oxford Univenity Press, 1959; Grove Press, New York, 1961) and D. S. Greenberg,
Tile Politics ofAmmcil" Sa,,", (penguin, London, 1969; first published IS Tile Poi,"s
of P", S""",, New American Library, New York, 196'7). I have quoted from them
occasionally to provide evidence for some particular points, but each in its own field
provides materials and insights complementary to those developed here.
2 Introduction
concepts in whose terms each of the problems could be discussed
in relation to the others.
I am well aware that my extended arguments can be difficult for
those who have not already begun to reflect on science in terms ofthis
new common-sense. In order to ease the way in, I have provided two
introductory chapters. The first shows that there is a problem in the
understanding of science. The second picks up several themes that
are current in popular discussions ofscience, and provides them with
a unified theme: the industrialization ofscience and its consequences.
In the main body of the text, the order in which topics are dis-
cussed is gov-emed mainly by the requirements of clearly defined
terms and established conclusions that are necessary for a proper
analysis at each point. I have therefore proceeded from the aspects
of science that are more abstracted from their social context, to-
wards those ofthe fullest complexity. Thus, it is possible to analyse
problem-solving in research with only an occasional reference to its
wider aspects, while the problems of quality-control and ethics in
science are difficult to appreciate without a systematic knowledge of
the craft ofresearch and the nature ofits products. Lest this ordering
from the simple to the complex give the impression that the earlier
topics are in any way more fundamental, I have referred forward,
informally, to later sections of the book whenever appropriate.
The sequence of major problems is then epistemology, sociology
and ethics, and applications in the technical and human context. On
the first, I am not so much concerned to refute the earlier assumption
that science somehow produces-truth by the application ofa standard
method, as to see how recognizeable, objective scientific knowledge
can (but need not) eventually emerge from the very personal and
fallible activity of research. On the second set of problems, I analyse
the sorts of social mechanisms and attitudes that are necessary in
order that worthwhile research may be done; and I also study the
conditions under which they are effective, and those under which
they are not. Having established the concepts required for a dis-
cussion of both the 'individual' and the 'social' aspects of scientific
work, I can enrich them further and discuss the applications of
'science' and 'the scientific method' both to disciplines concerned
with man rather than with nature, and to problems relating to tech-
nology and to human welfare. With all this accomplished, I can
return to the general problems ofthe present and future condition of
science, and offer a critical discussion of these. Because I build up
Introduction 3
my technical vocabulary as I proceed, using ordinary words in special
restricted senses, it is not easy to dip into the book halfway through
and follow the argument. This difficulty is unavoidable, but the
analytical index should help the reader locate the main discussion of
the technical terms that I use in a special sense.
There is one basic term that I have deliberately used in several
senses, and that is 'science' itself: A glossary ofthe various meanings
of the English word now in use would be lengthy and tedious; and if
one extended this to the German Wust1lSchll{t and the French
science positive, both of which made important contributions to our
inherited notions of 'science', then any worthwhile study would be a
major undertaking in itself. Also, the different parts of my analysis
apply to different (though overlapping) senses of 'science', and the
context has been sufficient for establishing, in a general way, which
sort is being discussed. Thus in Part I, the concern is with the natural
sciences, pure and applied; while in Parts II and III, the analysis
applies to any sort of disciplined inquiry. In Part IV, the 'social
sciences', pure and applied, are to be included as a major object of
attention; and in the Conclusion, my concern reverts primarily to
the natural sciences. I have tried to indicate these particular focuses
of reference by my use of words standing for 'science' in the text,
and by the examples discussed in the foomotes. Although this
method still leaves some obscurity, it seemed preferable to a pedantic
differentiation of terms.
Another ambiguity in usage should be pointed out, as it may
eventually prove to be a weak point in my sttueture of concepts. I
have not considered the problem of the different levels of aggrega-
tion of problems, methods and results, described loosely by such
terms as 'field', 'discipline', and 'science'. There was no place
in my analysis where I could perceive any differeRce being made by
the level ofaggregation assumed. Hence I have rather promiscuously
used these terms as synonyms; and the occurrence ofone rather than
another should not be taken as having some particular significance.
In almost all cases, I have kept the discussion of particular ex-
amples in foomotes separate from the text. Although this may be
inconvenient for the. reader, and also deprives the examples of some
of their force as evidence, there are several reasons for my adopting
this style. The consistent alternative, of providing evidence for every
thesis by a significant example from history or current practice,
thoroughly established as valid in itselfand woven into the argument,
4 Introduction
would have required a German style of scholarship, involving a
preparation time and .a book both several times longer than the
present work. Also since historians of science have themselves only
recendy come to share the new common-sense of science that I am
trying to formalize, reliable and app.ropriate case-studies are still
rare; and to provide comprehensive evidence for each of my theses,
I would have needed first to re-create much ofthe history ofscience.
There are other reasons for the separation of examples from argu-
ment, deriving from the character of the work itself. No set of
particular examples can provide conclusive evidence for general
theses of the sort argued here. Many of my arguments will appear as
restatements of the obvious to those involved in the practice of
science, and for them the examples will function as light relief. On
the other hand, those who find my conclusions repugnant or in-
comprehensible would be resistant even to the most persuasively
argued case-studies; they could well dismiss each of them as un-
representative instances. Hence my arguments will carry conviction
if and only if they strike the reader as offering real solutions to real
problems; and for this, examples that illustrate the argument with-
out entangling and lengthening it seem most appropriate.
Although the main intended function of this work is a practical
one, to assist in the understanding and solving of the deep problems
confronting science now and in the near future, I have consistendy
attempted to maintain a standard of rigour and consistency appro-
priate to philosophical discussion. Also, I believe that some of the
concepts and distinctions I have introduced in the discussion may
be of use in other general discussions of science. The basic concepts
fall into two classes, those concerned with knowledge and those con-
cerned with action. For the former, I define 'problem', and give a
sketch of its anatomy, distinguishing the different stages of develop-
ment of its materials, as 'data', 'information', 'evidence', and 'con-
elusion'. In analysing 'methods', I distinguish among the criteria on
which the controlling judgements of scientific inquiry are based:
particularly those of 'adequacy' and of 'value'. I make strong use of
the concept of 'pitfall', which more than any other explains how
apparently rigorous research can lead to disasters. Studying the
subsequent history of a solved problem, I distinguish between the
'result' itself, the 'fact' which may evolve from it (and I define this
term closely), and 'scientific knowledge'. The condition of 'im-
maturity' of a field can be defined in terms of the absence of 'facts',
Introduction S
resulting from the lack of those appropriate criteria of adequacy
which steer the investigation away from pitfalls.
Concerning action, I make strong use of distinctions within the
'final cause' of a task, such as the 'goal' which defines what is to be
attempted, the 'function' to be performed by the successfully
accomplished task, and the 'purpose' in human satisfaction to which
the performed function is a means. The operator on the task has his
own pwpose (or purposes) to be achieved, not necessarily identical
with the fulfilment of the goal; and for brevity I use 'motive' for his
reasons for being on that task at all. The application ofthis battery of
final causes to the social activity of science enables an analysis of the
problems of quality-control and ethics. Also, with it one can dis-
tinguish between the 'scientific problem' of the traditional sort, the
'technical problem' in which the goal is defined by the desired per-
formance of a pre-assigned function, and the 'practical problem'
defined by the achievement of given purposes. One can also use these
distinctions for an analysis of various pathologies of social activities
in general; in their terms I am able to define 'corruption',
Finally, all these concepts can be used to enrich the earlier dis-
cussion of important styles of scientific activity. The 'industrialized
science' of the present can be distinguished from the 'academic
science' that dominates the folk-memory of leading scientists of the
older generation, in terms of the capital-intensity of the tools of
scientific work and the consequent new social relations within the
world of science. We now see emerging a 'critical science', in which
science, technology, politics, and ultimately the philosophy ofnature
are involved, and which may be the most significant development in
the science of our age.
Part I
THE VARIETIES OF
SCIENTIFIC EXPERIENCE
INTRODUCTION
'WHAT IS SCIENCE?'
Vie11JS ofscience
Since natural science depends on the general public for its support,
in this period as never before, the public understanding of science is
crucial, in the long run, for the continued health of the community
of science. In many ways, the public appreciation of science leaves
nothing to be desired. We are all grateful for the comfort and
security of life that is achieved by modem technology, and prepared
to accept the claim that all these good things are the by-products of
scientific research. Moreover, in the common-sense understanding
of man's relations with his natural environment, and even with his
fellows, 'science' reigns supreme. Supernatural explanations of
natural events, even of great disasters, are no longer taken seriously.
The areas of ordinary life where inherited craft-wisdom is valued
more highly than the judgements of scientific experts, are shrinking
down to the vanishing point. This arises partly from the increasing
artificiality of our material environment, and from the rapid changes
in it as well as in our social environment. A sign of the triumph of
'science' is the reliance on textbooks for such personal crafts as
rearing children and even achieving a happy married life. This
tendency is more marked in the United States than anywhere else;
there the prevailing assumption is that every problem, personal and
social as well as natural and technical, should be amenable to
solution by the application of the appropriate science.
Without such a general appreciation of science, it would be
impossible for the scientific community to continue to grow, and to
lay claim to an ever-increasing share of the national wealth. But the
limitations of this appreciation must be understood if the support is
to continue on a genuine and healthy basis. First, what the public
appreciates is not the same as that for which the support is required.
In all the varied activities of science (including popularization,
teaching at all levels, technological application, and administration
and decision-making) it is research that lies at the nerve-centte.
'What;s Science?' 13
Without research-the continuous, disciplined advance from the
known into the unknown-these other activities would either lose
their meaning) or become stale, sterile and eventually corrupted. Yet
scientific research is most difficult, if not impossible, for a layman to
comprehend. The work itself cannot be appreciated without some
prior familiarity with the very specialized activity of investigating
problems in the context of an abstract, technical discipline; while
it has some similarities to the running of a business, or even to the
driving of a bus, its special character can be communicated only to a
layman with a well-developed imagination. And the object of the
research work, the new scientific result, is almost always so techni-
nical, so dependent for its meaning and significance on the context
of the existing information in the special field, that the layman can
never get more than the most general idea of what it is all about. In
this respect, the 'layman' may equally well be a scientist in another
field, for the problem of internal communication within the world
of science is certainly severe. But at least one working scientist
can recognize the moves in an exposition by another, in spite of
not understanding the content; while the man in the street would
find a detailed account of a new discovery to be totally incompre-
hensible.
Hence what the general public appreciates in science is not what
the scientists are doing. Rather, it can be classed under two headings:
techniques and natural magic. The first is the collection of devices
that make life easier to live, or the destruction of life more efficient.
The second is the production of strange and wonderful effects with-
out recourse to supernatural agents. Although the term 'natural
magic' fell into disuse some time ago, it is quite natural for this
interpretation ofscience to persist. Whet! the layman (child or adult)
sees a demonstration of some astonishing effect, can he imagine the
human endeavour and intricate technical work which led to its
creation? No; it is the effect itself which captures the imagination,
and produces wonder and delight: this is natural magic. Of course,
the audience for the 'wonders of science' is told that these achieve-
ments are simply the application of the laws of Nature, and that the
creator of the effect is in no sense a magician. But since the audience
cannot understand the laws ofNature that are relevant to the produc-
tion of the effect, it is strange and wonderful (until it becomes a
commonplace part of the environment) and so differs little from
natural magic of the old sort. We can see the strength of this attitude
14 The VII,ietieS of S'ientific Ezperitn&e
from the great popularity of projects that are politely described as
science but which are clearly recognizable as nearly pure natural
magic; the manned exploration of sPace is the best example.
In earlier ages, when natural science was pursued by a small band
ofenthusiasts with the support ofa handful ofpattons, the ignorance
ofthe general public did not usually matter. But it is now the general
public which, indirectly or directly, pays the piper, and which,
through its elected representatives, ever more frequently wishes to
call the tune. No amount of popularizing effort will overcome the
fact that at the present stage of civilization scientific research is a
rather esoteric activity. Even in their relations with their pattons in
the past scientists have sometimes been tempted to play something of
a confidence game, inventing a middle ground between what was
important to their pattons and what was worthwhile to them. The
temptation is even stronger now, but it must be resisted. If relations
between the scientists and the general public are to remain healthy,
they must be based on mutual respect between people in different
worlds. The scientist must not dismiss the layman as an ignoramus,
and the layman must ttust the scientist as a man whose work, while
incomprehensible, is genuinely worthwhile for its own sake. Such
mutual respect cannot be achieved until both sides recognize the
difficulties in the way of genuine communication; and of course
there is a special responsibility on the scientist to be worthy of the
trust placed in him.
Between the general public and the professional scientists lies a
rather large group, roughly between five and ten per cent of the
population, who have some detailed acquaintance with scientific
knowledge, although no experience ofresearch. These are the people
who have made a disciplined study of some branch of science;
students, present or past. It is widely hoped that a higher proportion
of places ofresponsibility in public life will be taken by such people,
so that there will be an intelligent and competent mediation between
the scientific community and the general public. In the student's
view ofscience there is assuredly little ofmagic; but unfortunately it
is faIse to suppose that the experience ofleaming a selection from the
body of existing scientific knowledge confers any insight into the
process of its creation.
Up until very recently the overwhelming impression of science
imparted to a student by all but a few eccenttic and gifted teachers
was that of a mass of accomplished, solid facts. The student's task
'What;s Science?' IS
was to assimilate a pre-sorted sample of these and to be capable of
reproducing them under examination conditions. The facts could be
of many types: special bits of information, techniques of manual or
mental manipulation, and general laws and theories. Whatever
their sort, they were recognizably hard, and announced to be clear.
For a student in such an environment, the frontiers ofscience would
be located on the remote heights of a pyramid of facts, and the work
of research would be imagined as the addition of some more hard
facts on top.
Much of the teaching of science in the later years of secondary
schools and in universities is still conducted within this framework.
But in recent years there has been a massive onslaught from the
proponents of the many reformed syllabuses for the teaching of
science. The difficulties encountered by these reforms may serve to
remind us that while this older view of science is dangerously one-
sided it is by no means without foundation. The great achievement
of the natural science of the past few centuries, unique in the history
of civilization, is the wealth of facts (of all the sorts I mentioned
above) which it has produced. And these facts are, to an astonishing
degree, hard and reliable, and are usually sufficiently clear to serve
their purpose. Moreover, scientific research, even of the most
inspired and revolutionary sort, is not accomplished by a great man
opening his eyes to the world about him, but necessarily grows out
of the matrix of a body of highly technical special results. Except in
the very rare cases when a whole field is created, it is on the basis of
old facts that new ones are made. Finally, to be able to do this work,
the scientist must be an accomplished craftsman; he must have
undergone a lengthy apprenticeship, learning how to do things with-
out being able to appreciate why they work.
Thus, the problem of imparting a correct view of science to
students cannot be solved by throwing out the dry old facts and
bringing in the excitement of open-ended research. Some sorts of
students may enjoy this much more, and those who go out into the
world will have a better appreciation of research than students of
earlier generations trained under the former discipline. But one can-
not be a craftsman unless one can manipulate one's tools; and one
cannot appreciate craftsmanship in others, as in judging- between
solid and spurious research, unless one has been trained up to it one-
self. Thus the students of the present day may have either one of two
diametrically opposed views of science, depending on whether they
16 The VII,ietieS ofScientific Experience
have been the subjects of pedagogical experimentation. Coming out
of the old school, they will tend to see science as the mountain of
facts; and coming out of the new school, they will tend to see it as
questions with ends opening in all directions. Some of the new
syllabuses manage to strike a balance; but in the difficult and
delicate task of training inexperienced and immature minds to
appreciate the complex interplay offacts and problems it is too much
to expect uniform success within a short time.
It might seem that the group of people whose view of science is
the most important and valid are the working research scientists
themselves. But although research is at the centte of scientific
activity, it remains as one very specialized part ofa large and complex
whole. The experience ofthe research worker will, in its own way, be
as specialized as that of the student. His task is to achieve new
results in a special field; even if he teaches part-time (as at a univer-
sity) there will usually be little or no connection between the new
results he is creating and the established knowledge which he is
passing on. Also, unless he is already in a position of seniority and
responsibility, he will have little involvement in the work which
requires him to see his own efforts in the context of the field as a
whole, in making the judgements and decisions which determine the
directions of future research.
Over the generations there has developed a view of science which
is well suited to produce the intensely committed, completely
specialized activity which modem science has required from the
majority of its practitioners. For them, 'research' is not merely the
centre of scientific activity; it is science. Even if this is not stated
explicitly, it can be inferred from the attitudes towards those
involved in other aspects of scientific work. Those who have tried
research, but abandoned it for easier things, are objects of some
contempt; those who have never even ttied, but are merely teachers,
are to be pitied.
This exclusive, sometimes fanatical, concentration on the produc-
tion of isolated scientific results can be traced back to the German
scientific research schools of the last century. It is conducive to the
production of solid work, but it is also anarchic and selfish. Its
prevalence has discouraged the growth to intellectual maturity of
many scientists who would otherwise have been capable of it. It
creates personal difficulties for scientists who, in their middle years,
would be more happily and productively employed on other sorts of
'What ;s Science l' 17
work related to science; but they are reluctant to have it said ofthem
that 'he has given up research'. The conditioning produced by years
ofconcentrated and narrow research has the effect ofmaking it more
difficult for the leaders of science to see the work of themselves and
their colleagues in its broader context, both within science and in
relation to the outside world. And finally, any social activity which
depends for its recruits on a supply of single-minded enthusiasts is
dangerously vulnerable to changes in intellectual fashion.
Science as the Pursuit ofTruth
The question 'What is Science?' supplies the title or the subject-
matter of many books on the 'philosophy of science'. In them the
question usually takes the form, 'What sort of truth is embodied in
completed scientific knowledge ?' Ideas developed in the course ofan
attempt to answer such a question will not be well suited for descri-
bing science as a human activity, always changing and never perfect.
Treatises on 'Scientific Method' written within such a framework of
ideas seem to have little relation to the real work of discovering new
knowledge and are frequently scorned by practising scientists who
have become amateur philosophers of science.
It would be a serious mistake to dismiss all this effort as irrelevant
to the understanding of science. The question itself is one of deep
and perennial philosophical problems, considered in the special con-
text of modem science. The essays in 'scientific method' are an
attempt to reveal the anatomy of scientific knowledge, analogous to
the way that Aristotle's logic of syllogisms functions for correct
verbal arguments. Moreover, many of the greatest 'philosophers of
science' have been scientists of distinction who were able to reflect in
a disciplined way on what they were doing. Even if this approach to
the question seems to provide few ofthe answers we are now seeking,
it will help our understanding if we appreciate why it once seemed
the best way, and also why it no longer does so.
The 'scientific revolution' of the seventeenth century, although
better described as a reformation in natural philosophy, had some of
the essential characteristics ofwhat we now recognize as revolutions.
In particular, it was inaugurated by a small group of prophets, fully
conscious of their role in attempting to destroy an existing order and
to replace it by something better. Their purpose was the achieve-
ment of truth, and they were committed to the view that this could
be achieved only through a certain approach to the study of the
18 The Va,ieties ofScientific Experience
natural world. A classic expression of this ideology can be found in
Galileo:
If this point of which we dispute w~e some point of law, or other part
of the studies called the humanities, wherein there is neither truth nor
falsehood, we might give sufficient credit to the acuteness of wit, readiness
of answers, and the greater accomplishment of writers, and hope that he
who is most proficient in these will make his reason more probable and
plausible. But the conclusions of natural science are true and necessary,
and the judgement of man has nothing to do with them. I
Although the men who consolidated and extended the scientific
revolution were generally neither so certain about achieving truth,
nor so exclusive in their claims for natural science, this Galilean
commibnent remained in the ttadition, available for renewed
emphasis in times of institutional and ideological struggle on behalf
ofnatural science. This occurred during the nineteenth century; and
a hint of its consequences for the self-consciousness of science, and
the philosophy of science, is given in the preface to the first edition
of Karl Pearson's classic Grammar of Science of 1892. There he
mentioned the fact that science previously had to carry on a 'difficult
warfare with metaphysics and dogma'; and the exigencies of this
struggle explained, for him, the deliberate concealment of the
'obscurity which envelops the principia of science'. 2 Pearson saw
that this 'warfare' had distorted the self-consciousness ofscience and
had harmed its teaching; and considering the battles as won, he
thought it appropriate to exhibit the structure ofscientific knowledge,
in its weaknesses as well as its sttengths.
The 'warfare' to which Pearson referred was documented by other
writers of his time; 3 and those who had fought for science were not
necessarily anti-philosophical or irreligious. But in Germany there
had been a bitter institutional struggle for the establishment of
I G. Galileo, DWope 011 Tile Gr,., WorM Systems, in The Salisbury Tnnslation, ed.
G. de SantillaDa (University ofQUcago Press, 1953), p. 63. The word 'judgement' in the
Jut IeDtence is • traDsJation of'arbitrio' and, like all key words, it is loaded with meanings.
The other modern English translation, that of S. Dnke, DWope COII&emi", The TJI)()
ClliefWtJrlll SystlllU (University of California Press, 1953), uses 'will' (p. 53). The latin
root bas the ambiguities from which the English derives both 'arbitrate' and 'arbitrary';
'judgement' seems best suited to this context.
:a Karl Pcanon, Tu Gr.""., of StinIe, (1St edn., 18g2; Everyman, London, 1937),
P·3·
J A. D. White,.A HisttJr:! oflu W.f.e ofStinlee ",;IA TlleoloO ill ClJristnulom, 2 vols.
(18g6; reprinIcd Do~ PublicatioDs Inc., 19(0), is the classic in this literature. White
was President of Comell University and American Ambassador to Russia.
'What;s Science?' 19
natural science as a discipline worthy of recognition as different
from, and independent of, the academic philosophy of the earlier
nineteenth century. And in England there had been a frequently
hysterical public debate over the theory of the evolution of the
human species, with ministers ofreligion denouncing the Darwinists
as the new Antichrist. What was at stake in these struggles was the
autonomy ofthe goals and methods ofinquiry in natural science; the
freedom of scientists to draw conclusions from the evidence they
considered relevant, and to ignore that which they considered
irrelevant (such as traditional Christian teaching, or accepted
metaphysical doctrine). In the last resort the struggle was over the
social functions and social responsibilities of natural science. Some
saw it as subordinate in its functions to some other sphere of
learning or work, in particular religion; and they considered that
scientists had a responsibility to work in such a way that both their
methods and their conclusions would reinforce the authority of
accepted religion rather than endanger it. The best defence against
this was, as usual, a good offensive; and the ideology of truth in
science, a truth independent from and perhaps superior to that of
philosophy or religion, performed a useful function in the struggle
for the freedom of science.
The struggle flared up briefly in the 1930s, when a group of
eminent British scientists (some native-born and others Continental
immigrants) saw a new threat to the freedom of science coming not
only from the racist persecutions of the Nazis but, more insidiously,
from the pressures on science in the cause ofsocial responsibility and
Marxist philosophy, enforced in the Soviet Union and urged by
British supporters of Soviet Communism. 4 The later episode of the
capture of Soviet genetics by T. D. Lysenko, who desttoyed a
brilliant school in the name of Marxism, seemed to confirm all their
fears. In reaction to this disaster, a distinguished American historian
of science used the Galileo text in a condemnation of the Marxist
approach to science and its history:
To suppose otherwise is to give the game away. It is to suppose that the
papal court was justified when it decided to condemn Galileo on the
ground that the Copernican system, however convenient as a mathematical
device, tended to unsettle men's minds and weaken authority and order.
It is to agree that the Soviet state was justified when it imposed
4 For a brief history of the 'Society for Freedom in Science' see Sir Solly Zuckerman
Scientists and W", (Hamish Hamilton, London, 1966), pp. 141 if:
20 The Varieties ofScientific Experience
Lysenkoism on the science of genetics as a theory comfortable to
Marxist dogma.... 5
Thus the Galilean commibnent to truth in science, seen as a neces-
sary means to the defence of its freedom, has continued down to our
own age, and will be expressed whenever the freedom of science
seems endangered.
Such a directly political motivation for the conception of science
as the pursuit of truth is only a part of a long and complex history.
Partly because of its indubitable successes natural science has
seemed for many decades to be the paradigm of genuine knowledge.
In a variety ofways, the pursuit ofscience, or a popularized image of
scientific work or scientific knowledge, has functioned as a religion
in substitution for, or in opposition to, the accepted religion of a
society. At the more practical level, the belief in the possibility of
attaining some truth that will live on after one's personal death, has
furnished a motive for selfless and dedicated work by many scientists;
and some schoolmasters, forced by circumstances to transmit what
is mainly a manipulative craft, derive self-respect and peace of mind
from the conviction that they are imparting clear and distinct truths
to their young charges. Since the pursuit of truth can appear to be
the most noble and harmless of human activities, this conception of
science helps to justify their work to scientists themselves; and it
also helps them argue for public support, especially when public
relations work is done on behalf of expensive experimental fields of
research.
The Technocratic Conception ofScience
The obsolescence of the conception of science as the pursuit of
truth results from several changes in the social activity of science.
First, the heavy warfare with 'theology and metaphysics' is over.
Although a few sharp skirmishes still occur, the attacks on the free-
dom of science from this quarter are no longer significant. This is
not so much because of the undoubted victory of science over its
ancient contenders as for the deeper reason that the conclusions of
natural science are no longer ideologically sensitive. What people,
either the masses or the educated, believe about the inanimate
universe or the biological aspects of humanity is not relevant to the
5 C. C. Gillispie, letter to Ameri"." Scientist, 45 (September 1957), 266A-'74A,
answering aiticisms of his review of J. Needham Scimc, lind Civiliulion in Chi".,
voL ii.
'What is Science l' 21
6 Science can thus be considered as one sector of the 'knowledge industry'. One study
of this (F. Machlup, The Production lind Distribution ofKno7IJledge in the United SllItes,
Princeton, 1962) estimates its product at 27 per cent of the G.N~P.; although his
definition of 'knowledge' would be better termed 'software'.
22 The VII,ietieS of Scientific Ezperitnce
to provide unlimited blessings for us all. Until very recently, there
was no systematic appreciation of the fact that as science grows and
penetrates industry it becomes industtialized. This is partly a matter
of scale; as the invesbnent in scientific research increases, both as a
whole and in the cost of individual projects, there arises a need for a
formal administrative machinery for decision-making and executive
action. Moreover, many practically useful scientific results can be
tteated as a sort of commodity, to be produced under contract,
tailored to the special needs of the purchaser. The bulk of the public
funds which are given to 'science' are in fact contracts for such
specialized products; and it is only natural for the older forms of
pattonage of science to be displaced by contractual agreements for
the production of results, whether directly useful or not. Hence, in
any large and vigorous field, the social atmosphere becomes increas-
ingly 'industtial' where a large organization, with labour force
directed to specialized tasks, produces the sorts of results for which
the directors have been able to obtain contracts from agencies which
invest in such production.
Without some such organization it would be impossible for the
scientific community of the present, and of the future, to operate.
But the assimi1ation ofthe production ofscientific results to the pro-
duction of material goods can be dangerous, and indeed destructive
of science itself. For producing worthwhile scientific knowledge is
quite different from producing an acceptable marketable commodity,
like soap. Scientific knowledge cannot be mass-produced by
machines tended by semi-skilled labour. Research is a craft activity,
of a very specialized and delicate sort. The minimum standards of
accuracy and reliability for worthwhile scientific results are extremely
high. But there are no automatic, external tests of its quality, neither
gauges to check its agreement with specifications, nor a market
mechanism for the public rejection of inferior products. Two sepa-
rate factors are necessary for the achievement of worthwhile
scientific results: a community of scholars with a shared knowledge
of the standards of quality appropriate for their work and a shared
commitment to enforce those standards by the informal sanctions
the community possesses; and individuals whose personal integrity
sets standards at least as high as those required by their community.
If either of these conditions is lacking-if there is a field which is
either too disorganized or too demoralized to enforce the appropriate
standards, or a group ofscientists nominally within the field who are
'What is Science l' 23
content to publish substandard work in substandard journals-then
bad work will be produced. This is but one of the ways in which
'morale' is an important component of scientific activity; and any
view of science which fails to recognize the special conditions
necessary for the maintenance of morale in science is bound to make
disastrous blunders in the planning of science.
It may appear that this concern is somewhat beside the point. It is
applied science, or technology, that enters into the system of indus-
trial production, and every national plan for science ensures that
'pure science' or 'basic research' should not be neglected. Unfortu-
nately, it is impossible to make a neat line of division between the
two sectors, allowing one to serve Truth and the other Caesar. It is
more than the simple fact that no piece of scientific knowledge can
be guaranteed 'pure', or free of application. In addition, much of
'pure' or 'basic' research involves capital outlay on an industrial and
occasionally gigantic scale, and those who undertake and organize
such projects must necessarily have some of the qualities of an
industrial manager.
No sector of science is immune from the problems of industriali-
zation; and, conversely, modem sophisticated technology is suffi-
ciently like pure science to share its delicacy, and is also prone to
special diseases of its own. For technological innovation is also
removed from the immediate controls of the market. Every impor-
tant new device or process takes several years of development before
it can be produced, and hence a decision to invest in development is
necessarily partly a speculative assessment of prices, demand, and
competition in the future. In military technology, the criteria for
decision are even more complex and speculative. Mistakes can be
made here, as in science; but here they cannot be quietly buried.
There is then a danger that a firm (or more commonly a government)
will try to redress an error by making it bigger, and all considerations
of what is economically viable (not to say socially desirable) will be
cast aside. Supersonic airliners are a case in point. Hence scientific
technology, as much as industrialized science, can suffer from a new
and dangerous form of corruption.
The 'Humanist' Critique
Now that 'dogma and metaphysics' no longer do battle with
science, the traditional opposition views are collected together under
the label 'the Arts'. This is usually considered to be an outsider's
24 The Varieties ofScientific Experience
view, seeing science as a new Leviathan; a body ofknowledge which
is esoteric, inhuman, and increasingly dominant. Those who hate
and fear this new science are even more mixed than the Oxford
classicists and Chelsea poets who provided the initial evidence for
C. P. Snow's famous lecture.? Indeed, such critics are even to be
found within the camp of science itself, especially among students
who have been discreetly press-ganged into a scientific career by
their schoolmasters, or research workers in the position of helpless
proletarians on the production-line of some modem research
establishment.
The debate between the 'arts' critics and the defenders of science
is necessarily confused since it depends on a partial understanding on
both sides, and also involves a mixture of personal philosophy,
politics, and quite directly professional concerns (as among the
teachers of upper forms of English grammar schools). Also, the
debate oscillates between two levels: that of scientific knowledge
itselfand that ofits technological applications. This further confuses
the debate, for while the results of scientific inquiry are, in their
statement, remote from specifically human concerns and values, the
industrial applications of those results are ttansforming the material
and social conditions of human life. A fully consistent 'humanist'
critic must not merely point out the sterility (in his terms) of know-
ledge of the natural world but he must also deny the human value of
modem scientific technology. This latter task can be done, but only
by fabricating a legend of some bygone 'good old days' when men
were poor but happy, or by denying the human importance of the
majority of the human race. The fully consistent defenders of
science also have their difficulties. It is easy enough for a certain sort
of intellect to find only ttiviality in the works of literature and
philosophy, and to claim that natural science has the only path to
truth. 8 But the social responsibilities of scientists, in the face of
, c. P. Snow, Tile T"", CtdMes lind lile S&imljjie Ref}(JIIIIitm, Rede Lecture, 1959
(Cambridge University Press, 1959).
• Thus, I. I. Rabi, 'Science and Technology', in Tile I",PIIt' ofSeimee tm Tec/mology,
eels. Warner, Morse, and Eichner (Columbia University Press, 1965), pp. 9-36, can say
'The arts are certainly central to life. Yet they are not the kind of thing that will inspire
men to push on to new heights •.• Even the works of Shakespeare, which are essentially
an exploration of human character, are really wonderful, glorified gossip. I am not
degrading gossip, because we live by it, but it does not take us outside ourselves, outside
the human race' (p. 20). In fairness to Professor Rabi, it should be added that his
c:omments on social aspecII of science, in his essay and in the subsequent discussion, are
• valuable IOUI'CC of information and insights.
'What is Science l' 2S
destructive and evil applications of their work, cannot be brushed
aside lightly. A tempting way out is to tteat the scientist's search for
truth about nature as an absolute good, and then to disclaim all
responsibility for the application of this knowledge at the hands of
the technologists. In an age when industtialized science and
science-based industry are so closely related this neat bifurcation
may strike an unsympathetic audience as shallow, if not indeed
hypocritical. 9
Between these extreme attitudes lies a rich mixture of sincere but
worried criticisms and defences. The critics of natural science
implicitly (or explicitly) contrast it to their idealized image of Arts
studies as explorations in the philosophy of man. The teaching of
science is claimed to be irrelevant to true education, and indeed
destructive of the values of education. It is authoritarian in style,
giving students no opportunity to develop their powers of criticism
and judgement. Dealing with an artificial and abstract subject-
matter, it gives the student no materials for developing himself as a
mature person in human society. Moreover, run the criticisms,
scientific knowledge cannot be part ofthe intellectual equipment ofa
cultivated person. Either one is a highly trained specialist, or one can
learn of science only at third-hand through populariutions. The
specialists themselves are accused of having a narrow vision of the
world since their training and professional practice removes them
from contact with genuinely human concerns. The increasing power
of natural science thus threatens the destruction of a humane
understanding among educated people as the humane studies are
increasingly deprived of prestige, of time in university teaching
programmes, and of resources for research. As a result, our thinking
about ourselves and the world around us becomes grossly material
and quantitative; the higher sensibilities and values are crushed
beneath the machine.
This comprehensive indictment would have more force if the
aggrieved humanists had previously achieved reliable methods for
teaching wisdom through their various approaches. But as Plato
9 L. Ie. Nash, in Tile NfI'"e of,lIe N.,,,.r Seim&es(Little Brown & Co., Boston, 1963),
distinguishes between the mom responsibility of the sdmrist, when he 'elects to take
put in the technologic exploitation of science for destructive purposes', and sdmrifie
fto7l1letlge, 'ethically as neutral as iron', capable of becoming swords or ploughshara.
He concludes, 'Ambivalence attaches to the works of science simply because their
technologic exploitations rest in the bands of men' (p. 113), but he does not discuss
whether a similar ambivalence could be justifiably attached to scientists.
26 The VII,ietieS ofScimtijiG Ezperimee
showed long ago, the task is not an easy one. Descartes' savage
indictment of his own humanistic education at one of the very best
schools of Europe should be read by all those who believe that the
virtues of 'humanism' are so obvious as to need no defending. lo
Moreover, most ofthe present criticisms ofnatural science can apply
to any developed discipline, be it classical philology or Shake-
spearean criticism. For illumination on the great, eternal questions
can be achieved and transmitted only by those rare individuals whose
intellect and vision are equal to the task. The vast majority of
scholars must do the humdrum work of investigating narrow,
technical problems, and teachers must try to impart some compe-
tence to immature minds. When people trained as research tech-
nicians are forced to be philosophers in their teaching, the results
can be disastrous. The deepest and noblest ideas of great minds,
when retailed by lesser men, can easily become banal commonplaces,
mockeries of the original inspiration. It is also worth remembering
that on occasion some 'humane' disciplines have suffered from
precisely that dehumanization of which natural science is now
accused; it was most marked in classics and history in Germany at
the end of the last century.11
In fact, the differences between 'Arts' and 'Science' disciplines
are more of degree than of kind. Rivalry and conflict between
different spheres oflearning is as old as learning itself. The perceived
lines of division, and the sharpness of conflict, are the result of
temporary and local circumstances. It can be very entertaining to
proceed from impressionist judgements on a very partial experience,
by undisciplined generaliution to erect abstract entities like 'Arts'
and 'Science', and then let this pair of straw men do battle with each
other. But without a recognition of the many-mdedness of the work
of the advancement of learning in any sphere, and of the systematic
difficulties common to all, little new understanding will be
achieved.
10 In the Diseows tle I. M,,1uHIe, 1st part, Descartes fint describes his enthusiasm for
the ftrious subjects in his hUlDlDistic curriculum, and then disposes of them one by one,
showing that they either confuse or mislead the mind, or aetua11y corrupt our morals, or
(u theology) are condemned to sterility by their own uscrtioDs. Only mathematics
survived his destructive analysis.
I I See F. Lilge, 1711 Ainue of u . .",,· IIIe F.lw, of IIIe Cmna U";""'sily (Mac-
Millan, New York, 1948), elL 4: 'Criticism and Satire of Academic Culture: N'1dZBChe'.
This unfortunate philolopher was not alone in his aitiqUCj the great Swiss historian
Jacob Burckhardt also stood aside from the industry of DUTOW &ctologic:al historical
scboIanhip (pp. 100-4)·
'Whllt ;s Science l' 27
It is worth recalling that the lecture which gave rise to the pro-
longed discussion of the 'Two Cultures' was really directed towards
quite a different problem. An earlier title was 'The Rich and the
Poor', and the purpose of the lecture was to condemn the ignorance
in both 'cultures' of a phenomenon which will determine our
future: the power of the new scientific technology for transforming
the lives of the men, women and children now helping to emerge
from what Snow calls 'the great anonymous sludge of history'. But
Snow's English audience was not prepared to comprehend the
challenge laid down in that thesis, and so the tea-cups rattled
agreeably while the revolution went on outside.
I.
any implications that they are concerned with man-power. They are not.
They are concerned with individual men.
This was written in 1958, and since then the calculations with
units of scientific manpower have increased steadily. Indeed, the
term 'scientist' itself is giving way to the acronym ~S.E. (Q!talified
Scientist or Engineer) as a grading for a particular class of skilled
manpower. If Sir Eric Ashby were a backward-looking romantic
then his worries could be dismissed; but he is a scientist of distinc-
tion, with a broad experience ofthe world ofscience and universities.
That his 'unforgiveable sin' has now become standard practice is
evidence of the complete victory of the technocratic conception of
science.
Science lives not by manpower units alone; and without some
ideal to replace that of the pursuit of truth it could soon degenerate.
Moreover, to the extent that science becomes organized around the
service of commercial and military industry, it will be subject to the
criticisms of being dirty work. Attempts by leaders of science to
conceal this connection from an audience of the lay public and
potential recruits, while using it to obtain support from industry and
the State, will accomplish nothing but a further decline in the
prestige and morale of science.
There is certainly no easy way out of this impasse; and to the
extent that industrialized science depends on large-scale investment
from industry and the State, it must be responsible to those institu-
tions. But it is possible that the reaction against runaway technology
will not take a purely negative form. Many leading scientists have
14 Eric Ashby, Te&1mo101] .tUl tile A&fIdemi&s (MacMillan, London, 1958), p. 94-
30 The Varieties ofScientific Ezperimce
already begun to criticize what is done in the name of science; and
there are now emerging focal points of 'critical science' where
research is organized around the identification and exposure of the
damage being done to the biosphere. As yet this movement is small;
and, even if it grows, it will necessarily exist on the margin of
industtialized science. Also, it must expect incomprehension and
hostility when it tries to establish an institutional base for its
teaching; for by its programme it implicidy rejects our inherited
approach to science, with its faith in Facts and in Progress, as a relic
of the Victorian age. But as it gains strength and coherence it will
develop a new philosophy of science, and a new philosophy of
nature and ofman's place in it. For this, it can draw on a suppressed
tradition within natural science itself, which saw beyond the
accumulation of facts, and beyond the domination of nature, to the
welfare ofhumanity living in harmony with itselfand its neighbours.
Whether such a philosophy could flourish within the context of our
industrial civilization, and whether the new science based on such
a philosophy could gain influence in time to avert the destruction of
civilization, are unanswerable questions. But it is certain that without
a new understanding of science our present problems, within
science and in its applications, will multiply to the point ofno return.
2
SOCIAL PROBLEMS OF
INDUSTRIALIZED SCIENCE
IN recent years, the vision of'science' as the pursuit ofthe Good and
the True has become seriously clouded, and social and ethical
problems have accumulated from all directions. Each particular one
of these can be discussed in isolation, and solutions sought as if it
were the only or the main perturbing force in an otherwise tranquil
scene. But there are systematic connections between these various
problems, which can be organized under the tide 'industtialization
of science'. This means, in the first place, the dominance of capital-
intensive research, and its social consequences in the concenttation of
power in a small section of the community. It also involves the
interpenetration of science and industry, with the loss of boundaries
which enabled different styles of work, with their appropriate codes
of behaviour and ideals, to coexist. Further, it implies a large size,
both in particular units and in the aggregate, with the consequent
loss ofnetworks ofinformal, personal contacts binding a community.
Finally, it brings into science the instability and sense of rapid but
uncontrolled change, characteristic of the world of industry and
trade in our civilization.
The social and ethical problems consequent upon the industriali-
zation of science are the deepest problems in the understanding of
the science of our period. In response to these problems, both the
informal common-sense understanding of science, and the scholarly
philosophy of science, have been shifting their perspectives and
modifying their assumptions about what is 'real' in science. The
present work is intended as a contribution towards the under-
standing of these problems; and before going into a lengthy discus-
sion ofthe foundations, it will be useful for us to review the problems
themselves. Much of what I say will be harsh and unpleasant, and
to some will appear to be muckraking for its own sake. But these
32 The Varieties of Scientific Ezperimee
things can and should be said, not only because they reflect the
judgements of eminent scientists, but also because there are many
men who have devoted their lives to science, whose efforts to solve
these problems will be helped by their identification and explana-
tion.
Paradise Lost
A very important feature of the social problems of contemporary
science is their novelty.. Although the gross quantitative indices
of scientific work show a fairly steady growth-rate over many
decades,1 the self-awareness ofthe scientific community, as reflected
in the pronouncements of leading men in leading fields and in the
sort8 of studies made of science, has experienced a sudden change.
Whether the situation of science as a whole has really altered so
drastically in the very recent period is not of crucial importance in
this context. For the belief in the rapidity of change means that the
present leadership of science considers itself as lacking a stock of
inherited attitudes and social skills for coping with the problems
created by the present condition of science. As we shall see, the new
awareness of social and ethical problems of great complexity,
together with a sense of changes beyond the control of the scientific
community itself, and of inexperience in handling them, are them-
selves factors in a complex and unstable situation.
The sense of loss of a very precious quality of pre-war science is
conveyed clearly in the reflections of the distinguished Soviet
physicist Kapitsa:
The year that Rutherford died (1938) there disappeared forever the
happy days offree scientific work which gave us such delight in our youth.
Science has lost her freedom. Science has become a productive force. She
has become rich but she has become enslaved and part of her is veiled in
secrecy. I do not know whether Rutherford would continue nowadays to
joke and laugh as he used to do. 2
The loss of innocence of science is even deeper than Kapitsa
indicated. Science is not merely harnessed to industry; it can be
I The pioneering work in estab6ahing these statistical regularities and in appreciating
their sipificance was done by D. ]. de S. Price. See his ullie Snmee, Bi, StinI&e
(Columbia University Press, 1963). I was informed by the author that the basic quantita-
tive regularities were discovered by the techniques of 'little science': working from data
obtained by IUDpliDg IlaDdard ~ works. The computers were brought in later.
2 Qpotecl from v. K. McElheny, 'Kapitsa'. VISit to England', Snmee, 153 (1966),
725-7·
Social Problems ofIndustrialiutl Science 33
applied to produce effects which for some centuries, since the
decline ofbelief in magic and miracles, had been thought impossible.
Norbert Wiener described this revival of ancient moral problems
under the heading of simony: the perversion of the magic of the
Host to other ends than the Greater Glory of God.
Dante indeed places it among the worst of sins, and consigns to the
bottom of his Hell some of the most notorious practitioners of simony of
his own times. However, simony was a besetting sin ofthe highly ecclesias-
tical world in which Dante lived, and is of course extinct in the more
rationalistic and rational world of the present day.
It is extinct I It is extinct. It is extinct? Perhaps the powers ofthe age of
the machine are not truly supernatural, but at least they seem beyond the
ordinary course of nature to the man in the street. Perhaps we no longer
interpret our duty as obliging us to devote these great powers to the
greater glory of God, but it still seems improper to us to devote them to
vain or selfish purposes. There is a sin, which consists of using the magic
of modem automatization to further personal profit or let loose the
apocalyptic terrors ofnuclear warfare. If this sin is to have a name, let it be
Simony or Sorcery.3
Still more serious is the changed perception of the character of
scientific work, and of the people who engage in it. For in the
traditions of propaganda on behalf of science, as well as in the
practice of the social activity of scientific inquiry, the work requires
codes of behaviour which are more sophisticated and more enlight-
ened than those adequate in many other spheres, and hence the
scientific community requires members ofsufficient moral calibre to
adhere to them. This ethical aspect of science has long been an
important part of its self-awareness. Since science cannot honestly
be defended purely on its contribution to industry, and its products
are generally too esoteric to contribute to general culture even at the
highest level, the work of scientific inquiry as a worthwhile thing in
itself must be an important component of any justification of the
total activity of science.
Traditionally, the image of science has been of a work which is
demanding, and also productive, of the highest standards of
morality. The work is arduous, involving the sacrifice of leisure and
convenience and lacking any of the rewards of wealth and power
which are supposed to spur ordinary mortals on in their endeavours.
In addition, scientists offer the products of their work openly and
3 N. Wiener, Coil tS Colnn, 111&. (M.I.T. Press, 1964), pp. 51-2.
34 The V.ties ofScientific Ezperim&e
without charge to all colleagues, ignoring the barriers of politics,
nationality, race, or class. Moreover, the inevitable debates over
scientific results are conducted within a set of rules which seem
inhumanly strict, and yet which are adhered to by consensus. The
scientist cannot hire an advocate to make his case sound better than
it is; he must not resort to personal abuse or denigration of oppo-
nents, or use his political influence (within the community or with
lay society) to determine an issue; and any resort to dishonest
reporting ofwork is sufficient to ruin a man's reputation. The willing
adoption of such rigoroUS standards of practice, and their operation
in the absence of any formal institutions for enforcement, seems to
require individuals of a genuinely superior moral character, who
willingly submit to such a discipline in the service of a noble end:
namely, the advance ofknowledge and power on behalfofhumanity.
There is no doubt that this elevated view of the ethical aspects of
science has had a strong foundation in reality, in the social practice
of science as well as in its public self-awareness, during the period
of'academic science' which began to be displaced during the Second
World War. The initial contact of scientists raised in pure academic
science, with men of affairs, often produced reactions like the one
expressed by the physicist Szilard:
When a scientist says something, his colleagues must ask themselves
only whether it is true. When a politician says something, his colleagues
must first of all ask, 'Why does he say it?'; later on they mayor may not
get around to asking whether it happens to be true. 4
The context of this quotation is a story called 'The Voice of the
Dolphins', in which an international organization of scientists has
found the simple secret of gaining the adherence of any politician to
any cause, irrespective ofhis predispositions: find his price, and buy
him. Szilard's estimated round figure at which any American
politician's soul could be purchased was a guaranteed lifetime
annual income ofS2oo,ooo (at 1960 prices).5
In recent years, the ttaditional portrait of the noble scientist has
receded from the centte of the image of science. There has not been
a denial of its accuracy for that part of scientific work which it
describes; but there has been a growing realization that there is more
to the activity of science than the achievement and assessment of the
4 See L Szilard, TM Voil, of 1M Doll_'" IIIIIl 01. Slories (Simon & SchUlter,
New York, 1961), pp. 35~.
5 Ibicl., p. 41, where the point is illustrated by example.
Social Problems ofIndustrialiutl S'ience 3S
results of research. And the tangible rewards of a successful career
in high-prestige pure science, and in industrialized science, are
common knowledge. Szilard himself recognized that when scientists
engage in work other than straightforward research they are subject
to the same hazards as anyone else involved in practical affairs. In
another short story, he sketched a plan for achieving a desired
deceleration of scientific progress in the future through the creation
of an administrative machinery that would turn scientists into
bureaucrats. In this way, talented men would be diverted from
research, and mediocrity would be systematically fostered. His
model for this stifling bureaucracy was, very unfairly, the American
National Science Foundation. 6
Nostalgia for the bygone simplicities of particular fields of
science can exaggerate the depth and novelty of the social problems
ofcontemporary science, and thereby hinder their solution. Also, the
confusion among lay audiences between 'pure' and 'applied' science
and 'technology' obscures real differences which must be appreciated
if the different aspects of scientific work are to be understood and
maintained in a healthy relation. Yet this confusion is not altogether
unjustified; the creation of nuclear weapons was the achievement of
men who had been trained in the most pure and philosophical field
of science, atomic physics. And even 'pure science' has come to
require gigantic projects such as high-energy particle accelerators,
involving vast funds and also managerial and political skills for their
achievement.
In these new conditions, there has emerged a new literature about
scientists, in which the scientist is seen primarily as a man of affairs,
and the painstaking, disciplined work of research is only a part,
perhaps a minor part, of the story. There have been protests at this
practice of ignoring the science when talking about scientists;7 but
6 0p. at., pp. 89-102, 'The Mark Gable Foundation', especially pp. 100-1. The plan
was to have the best men in each field appointed as high-salaried chairmen, and to have
about twenty annual prizes of 1100,000 for the best papers of the year. 'As a matter of
fact, any ofthe National Science Foundation bills which were introduced in the 79th and
80th Congresses would perfectly wen serve as a model' The effects would be achieved by
taking the best scientists out of their laboratories, and inducing the others to follow
fashions.
7 See N. Reingold'S review of Or,a";zat'01I for E,tJtUJmi, CO-Ojlerat'on au Dewlopmmt
(D.E.C.D., Paris, 1968), and of E. B. Skolnikoff, Stimee, Te'/moloo .rul Amni,."
Forn", Pol',y (M.I.T. Press, 196'7), and of D. S. Greenbtrg, The Politi" of P",e
Stimee (New American Libnry, New York, 196'7), Isis, S9 (1968), 216-20. Against the
first two, he argues that 'one cannot understand either (social and economic) output or
process without detailed knowledge and understanding of the 'input', which consists of
36 The Varieties ofScientific Experience
it appears that for the discussion of the social problems of science at
a reasonable depth, it can be done. Also, men of administrative
experience have (incorrectly, in my view) dismissed the ethical
component of scientific research as a purely prudential strategy by
scientists with application only to an isolated part of their work. 8
Even more corrosive ofthe ttaditional image of the scientist than
all this explicit discussion are the implicit assumptions that underlie
all debates on 'science policy'. For it is assumed that, when decisions
are made at the highest level on priorities between projects and
entire fields, the work for which support is offered will attract
scientists to it, regardless of their prior interests or views on the
choice. Thus, the bulk of the scientific community can be and is
treated as manpower, which will flow as directed between tasks for
which it has the requisite skills. As we shall see, such policy decisions,
and the resultant flows, are necessary in the conditions of indus-
ttialized science; and the enterprise could not survive if the majority
of competent scientists were extreme individualists who would work
well only on problems of their own choosing. But such manpower-
units cannot be considered as scholars, and any propaganda that
projects the image ofthe typical scientist as an independent searcher,
following his own path in the exploration of Nature, is now worse
than false: it is a bore.
Whether there has been a Fall of Scientific Man over the last
generation is a topic on which any debate will produce more heat
than light. The world ofscience is variegated in the extreme, and our
knowledge of the previous period depends largely on a folk-history
of science which naturally stressed certain aspects of science and
ignored others, and created a simplified, idealized picture out of
disparate and contradictory elements. But we have already seen that
the industrialization of science brings with it certain strong tenden-
specific scientists working in particular scientific contexts'. Greenberg's book he considers
to be 'not about pure science or even about pure scientists', but rather about power,
intematioDal hostility, industrial competition, and regional conflicts; the real focus of
the 'politics of pure science' is not in the scientific establishment but in the State and
industrial agencies that wield real power. In general, R.eingold observes that the one-
sidedness of such books reflects the division and weakness within the history of science,
where 'intemalist' and 'extemalist' approaches, although DO longer tied to hostile
ideologies, make hardly any contact.
• D. S. Greenberg quotes the plain-cpoken Robert M. Hutchins, 'There have been
very few scientific muds. This is becaUle a scientist would be a fool to commit a scientific
mud when he am commit muds every day on his wife, his associates, the president of
the university, and the grocer•••.' (The Polities ofP.t Slimet, pp. Zo-I.)
Social Problems ofIndustrialiutl Science 37
cies to change, in the realities ofthe social practice ofscience as well
as in its self-awareness. In this chapter I will describe some of these
new problems, and briefly indicate some of their causes. Because
many of the phenomena themselves are ruled out as inconceivable,
let alone inexplicable, in the terms of the traditional philosophies of
science (in particular, the circumstances in which scientific inquiry
proves abortive), it will be necessary first to establish the elements of
a new philosophy of science for their systematic description and
analysis. On that basis we will be able to return to a deeper study of
the social problems of industrialized science, and consider the ways
in which they might possibly be resolved.
The World ofAcademic Science
It will be useful for our purposes to sketch some aspects of the
social activity of science in the period preceding the present one. 9
The institutions and attitudes of science and scientists at the present
time are largely inherited from that period, and the memory of that
era is a point of reference for all analyses of the present except for
those which see science simply as a factor of production. This
period may be considered as having its terminal points in the French
Revolution and in the atomic bomb. The event that marks its
inception is the establishment of the Ecole Polytechnique in Paris,
where scholarship students chosen by ability received a scientific
education from the great masters. Although the school was designed
to provide engineers, and only a small proportion of its graduates
were able to pursue a .scientific career, it provided a model and
inspiration for later scientific education, as distinct from practical
training in medicine or engineering. Within a few decades, first in
Germany and then more gradually in England, science achieved an
institutional base in teaching posts in universities, which themselves
were reformed from their previous stagnation. All the other social
institutions internal to science were either created during this period,
or changed radically from their earlier forms. Generally, dilettantes
and a lay audience were gradually excluded. The work became con-
centrated and specialized, with an apparatus ofjournals and specialist
societies, and an increasing dependence on an occupational base in
teaching and research. The work of science itself progressed at an
9 A most useful collection of papers on the institutional aspects of contemporary
science, seen in the perspective of the recent past, is Sdm&t.rul Sotiet~, eel. N. Kaplan
(Rand McNally, Chicago, 1965).
38 The Varieties ofScientific Experience
accelerating pace; a succession of new fields were opened up for
effective disciplined inquiry, and the increase in factual knowledge
was stupendous. By the later part of the century the centre of
organized research was in the German universities, and there the
system of academic science may be said to have achieved its greatest
moments.
By the end ofthe nineteenth century there were two developments
that in rettospect can be seen to mark the inception of the latter
phase of the cycle of development of academic science as a social
institution existing in a particular, historically conditioned, context.
First, the German universities ceased to expand at their previous
rate; and second, the systematic institutionalized connection of an
important science with industry was established. This·was chemistry,
which had always had a very close relation with practice, but which
in later nineteenth century Germany became the basis for a sophis-
ticated, capital-intensive industry on the pattern which is now
dominant. The example and ideals of academic science, on a model
derived from Germany, remained vigorous well into the current
century. 'Science' has been understood as pure, university-based
science, in spite ofan involvement ofscience in the First World War
so deep that it has been called 'the chemists' war'. Indeed, not until
the Second World War had produced a scientific-technological effort
of a new order of magnitude, culminating in the atomic bomb, did
the interpenetration of industry and science, and the resulting
industtialization ofscience, destroy the claim ofthe 'academic' image
to represent the essential nature of science.
Our description ofthe social practice and self-awareness ofscience
in this past era will be under- four headings: the assurance of a
diffuse social benefit; the ethic of the search for truth; work in the
context of an autonomous community of gentlemen; and the par-
ticularly refined sort of personal property achieved in the work. For
the first two themes, we can do no better than to quote from the
classic pronouncement of Helmholtz:
In fact, men ofscience form, as it were, an organized army labouring on
behalf of the whole nation, and generally under its direction and at its
expense, to augment the stock ofsuch knowledge as may serve to promote
industrial enterprise, to adorn life, to improve political and social relations,
and to further the moral development of individual citizens. After the
immediate practical results of their work we forbear to inquire; that we
leave to the uninstructed. We are convinced that whatever contributes to
Social Problems ofIndustrialiutl Science 39
the knowledge of the forces of nature or the powers of the human mind
is worth cherishing, and may, in its own due time, bear practical fruit,
very often where we should least have expected it. Who, when Galvani
touched the muscles of a frog with different metals, and noticed their
contraction, could have dreamt that eighty years afterwards, in virtue of
the self-same process, whose earliest manifestations attracted his attention
in his anatomical researches, all Europe would be traversed with wires,
flashing intelligence from Madrid to St. Petersburg with the speed of
lightening? In the hands of Galvani, and at first even in Volta's, electrical
currents were phenomena capable ofexerting only the feeblest forces, and
could not be detected except by the most delicate apparatus. Had they
been neglected, on the ground that the investigation ofthem promised no
immediate practical result, we should now be ignorant of the most
important and most interesting of the links between the various forces of
nature....
Whoever, in the pursuit of science, seeks after immediate practical
utility, may generally rest assured that he will seek in vain. All that science
can achieve is a perfect knowledge and a perfect understanding of the
action of natural and moral forces. Each individual student must be con-
tent to find his reward in rejoicing over new discoveries, as over new
victories of mind over reluctant matter, or in enjoying the aesthetic beauty
of a well-ordered field of knowledge, where the connection and the
filiation of every detail is clear to the mind, he must rest satisfied with the
consciousness that he too has contributed something to the increasing
fund of knowledge on which the dominion of man over all the forces
hostile to intelligence reposes.
II I am indebted to Mr. N. Parry for forcing me to realize the necessity of this par-
ticular argument.
Social Problems ofIntlustrialiutl Science 43
astronomy was always capital-intensive and related to the State. IZ
But the simplified picture on which my analysis is based does have a
certain historical reality: it was the dominant self-consciousness of
academic physical science in Germany in the latter part of the nine-
teenth century, whence it spread to other fields and other nations. 13
As such it is important, for it is such an ideology that provides the
only existing foundation for an ethics of science which is distinct
from that of technological development or commerce. 14 The danger
is that this ideology will be kept in a fossilized state for particular
public-relations functions, while becoming less and less relevant to
the experience of those who live in the world of industrialized
science. Then the inherited ethical principles of scientific work will
12 The basic historical studies of the social and institutional aspects of European
academic science have been made by Joseph Ben-David. His earlier papers are listed in
Scimee flfIIl Society, ed. N. Kaplan (Rand McNally, Chicago, 1965), pp. 581-3. A
pioneering dort to characterize the difference between 'academic science' of the nine-
teenth century and the earlier socW forms, is his 'Scientific Growth: A Sociological
View', Milltl"Wl, 2 (1964), 4SS-T1.
13 In 'The Scientific Worker' (ph.D. thesis, University oC Leeds, 1969), N. D. Ellis
has explored some of the ambiguities in the concept of the 'autonomy' of academic
science. He suggests that the legend oC 'pure science' may be a confIation of two quite
different situations: the German academic, 'pure' with respect to applications, but unless
a full ProCessor, nol autonomous in his choice of problems; and the British amateur, free
to Collow his choice, but not at all averse to involvement in commercial ventures. More
detailed historical studies are required before this insight can be considered as fully
established.
14 Two examples, widely sepanted in context, indicate the great variety of principles
that can be invoked Cor the preservation of the integrity of science. An extreme case is
cited by D. S. Greenberg, Tile Politits ofAmerit_ Sdmee (p. 93), concerning a group of
distinguished biologists in the early twentieth century, who preferred penury Cor them-
selves and for their work, rather than accepting grants from the Carnegie Foundation and
from the University ofChicago which had 'unacceptable' conditions attached. A sttiking
contrast is provided by a comment on a priority dispute between Charles Wheatstone,
F.R.S., and then ProCessor at King's College, London, and an obscure mechanic,
Alexander Bain. In earlier nineteenth-century London, it was not thought wrong for a
man of science to concern himself with pnctical (and profitable) matters; but the
distinction between the two sorts of work was clear to this public. In this case, the
dispute was over a patent, and 80 the reviewer of. book on the controversy could write:
'The ProCessor, in the case before us, has laid aside his gown-he has stepped from his
chair into the shop oC the artisan. His discoveries there are not the treasures presented to
the scientific world, they are riches heaped up Cor his own private emolument ••• when
a philosopher avails himself oC the protection of the Patent Laws, he losts tllSle as a
philosopher, and descends to an equal rank with all others who seek the protection of the
same banner.' The quotation is from a review of J. FinIaisson, A" Atttnmt of SOffIt
Rt1IIMill1Jle Applitflt;oru oftile Elft,"t Fluid 10 lilt UstfiJ bts by M,. AleztllUler B." •.. ,
in the Eltaritfll Mflg.R;ine (October 1843), pp. 139-42. I am indebted to Mr.1l. A. Muir
Cor this reference. An account of Wheatstone's career, with a sympathetic history of his
involvement in another priority dispute, is G. Hubbard, eDDie flnil W1IefltStone.nil tilt
Int~lio" ofllle Eltttrit Telegraph (Roudedge, London, 1965).
44 The Varieties ofScientific Ezperience
become increasingly divorced from reality; and under the pressures
of present conditions, they could not long survive as effective
conttols on action.
The Industrialiution ofProduction in Science
For a comparison of the present age with the one preceding it, we
can best start with the changed technical character of the work of
scientific research; for from this follow the changes in its social
institutions, and social practices. ls The basic difference is a simple
one: research is now capital-intensive. Any significant piece of work
is almost certain to cost far more than an individual scientist can
afford out of his own pocket; it will generally cost much more than
his annual income. Hence he is no longer an independent agent,
free to investigate whatever problem he thinks best. Nor is he likely
to have personal contact with a private patron who will provide for
all his needs. Rather, in order to do any research at all, he must first
apply to the institutions or agencies that distribute funds for this
purpose; and only if one of them considers the project worth the
investment can he proceed.
This change is as radical as that which occurred in the productive
economy when independent artisan producers were displaced by
capital-intensive factory production employing hired labour. The
social consequences of the Industrial Revolution were very deep,
and those of the present change in science, while not comparable in
detail, will be equally so. With his loss ofindependence, the scientist
falls into one of three roles: either an employee, working under the
conttol of a superior; or an individual outworker for investing
agencies, existing on a succession of small grants; or he may be a
contractor, managing a unit or an establishment which produces
research on a large scale by contract with agencies. Of course, he may
have other tasks and responsibilities, as refereeing for journals,
IS I am particularly indebted to Professor D. Humphrey, of Oregon State University,
for helping me develop the ideas of this chapter. I should also say that even before I
bepn to think seriously on these problems, they had been indicated by John Ziman in a
broadcast ta1k, 'Scientists, Gentlemen or Playas?', Tile Listener, 68 (1960), sw-607.
For surveys of the history of the process whereby academic science evolved into industrial
lCience, see D. S. L. Cardwell, Tile Or,fI";UtUm of StinIe, ill £",111l1li (Heinemann,
London, 19S7), and H. and S. Rose, StinIe, fIIUl SodIl;y (Allen Lane, London, 1969).
It is interesting that throughout their historical study, the authors of the latter book
abow impatience with the slow growth of industrialization over the past century; but in
their final chapter they find themselves making a ndical aiticism of the state of affairs
that baa fiDaIly been ad1icvecL
Social Problems ofIndustrialized Science 4S
advising investment agencies, and so on; but in relation to his means
of production and the decisions which determine his work, his
position must be one of these three.
Along with this differentiation of the positions of individual
scientists, there comes a concentration of the power to make
decisions, and the development of a formal administrative system
for this function. The dispersal of large sums of money, and even
more the decisions between competing demands, are matters which
require proper procedures of information and control. A completely
informal consensus of a large community is not sufficiendy precise
or reliable to be the basis for such work; and the investing agencies
must work from the evaluations and judgements of a group of
advisers. With this concentration of powers of decision and control,
the free market place of scientific results, whose value is established
after they are offered and by an informal consensus, is replaced by an
oligopoly of investing agencies, whose prior decisions determine
what will eventually come on to the market.
This is an inevitable consequence of the costs of research; even if
individual universities were given large block grants out of which to
finance all the research done by their members, the same sorts of
decisions would be necessary, and a similar set of structures would
develop.16 Also, the investing agencies try to maintain some of the
old conditions of consensus, by choosing their advisers from among
the recognized leaders of each field. But the incoming generation of
leaders, and all those following, have built their careers and reputa-
tions under these new conditions, and will be subject to its influences
in ways which I shall soon describe.
The most basic effect of these technical and structural changes is
a tendency to a change in the location ofthe intellectual property ofa
scientist, possession of which is desired for the achievement of his
personal purposes. We recall that under the old system it was
fundamentally the published research report that constituted his
property; on the basis of the informal evaluations of it by his
16 In his broadcast talk on 'Planned Science' in 1948, Michael Polanyi argued mgendy
that 'no committee could forecast the routine progress of science except for the routine
extension of the existing system'. Hence, 'The function of public authorities is not to
plan science, but only to provide opportunities for its pursuit. All that they have to do is
to provide facilities for every good scientist to follow his interests in science.' Polanyi
may have been correct in believing that the Marxist impulse to the planning of science
had evaporated; but the impossibility of carrying out his suggested policy soon made
planning inevitable. See The Logie of LilJerty (Roudedge, London, 1951), pp. 86-90.
46 The Varieties ofScientific Ezperimce
colleagues, he expected appropriate rewards in his career, and the
personal satisfaction produced by public recognition of his work. In
the present situation, the research contract is not merely a pre-
requisite for the future possession of the property embodied in a
published paper; it also brings immediate benefits in itself, in the
way of prestige and possible material conveniences. Moreover, with
the concentration of decision-making power to the investment
agencies and their few advisers in each field, their estimate of a man
is of more practical significance for his career, than that of some
future diffuse consensus. Hence the location ofa successful scientist's
property tends to shift from his published results to his existing
research contracts, and the personal contacts that will ensure their
continuation. With this shift in the location of the scientist's
property, there is a tendency to a corresponding shift in his concep-
tion of a successful career. Especially for someone who enjoys the
role of a contractor, with its incidental benefits, the goals of a
career in science can change from being a series of successful
research projects made possible by a parallel series of adequate
contracts, to being a series of successful research contracts made
possible by a parallel series of adequate projects. I? When this
happens, the man is better described as a 'scientific entrepreneur'
than as a 'scientist'.
Under such conditions of division of labour, concentration of
decision-making power, and tendency to the shifting of goals, it is
impossible for a 'community' to survive in its old form. The mere
expansion in size of every field, speciality, and subspeciality, would
make it difficult in any case. But under the present conditions,
differences in wealth (taken here as a measure of the scientist's
intellectual property) produce such extreme differences in prestige,
power, and material benefit, that we can ttuly speak of classes in a
society ofscience, rather than of more and less eminent colleagues in
a community. Since most scientists engaged on 'pure' research are
employed by a university (or by one of the few State-supported
research laboratories where conditions are equivalent to those at
universities), they will tend to settle at institutions of a prestige
status comparable with their own. The leading men, forming a very
perceptible 'invisible college', will congregate at the great univer-
.7 This distinction is made implicidy by D. S. Greenberg when he contrasts 'grants-
manship', the (ethically neutral) practice ofextracting funds for research, and 'chiselling'
(Tite Politi&s of Alllerita Stim&'t pp. 351-3).
Social Problems ofIndustrialized Science 47
sities, enjoying favourable conditions of employment there, and with
their intimate connections with the investing agencies (a prerequisite
for the existence of such a group) they will pursue the researches
they please in comfort. And since in most countries the high-
prestige universities cluster in a special geographical region, those
who are left, or cast, out of the charmed circle of the successful men
may be isolated in every way. Deprived of the personal contacts
whereby one keeps up with new work, and whereby one makes an
impression on those who advise the investing agencies, they are left
to do derivative or second-class research on less generous grants, or
none at all.
Would it be too much to say that in the natural sciences today the given
social environment has made it very easy for even an emotionally unstable
person to be exact and impartial in his laboratory? The traditions he
inherits, his instruments, the high degree of specialization, the crowd of
witnesses that surrounds him, so to speak (if he publishes his results)-
these all exert pressures that make impartiality on matters of his science
almost automatic. Let him deviate from the rigorous role of impartial
experimenter or observer at his peril; he knows all too well what a fool
So-and-so made of himself by blindly sticking to a set of observations
22 The American Psychological Association baa recendy found itself in mnfliet over a
project designed to provide an institutional machinery for the publication of untested
materials. An 'early dissemination' scheme, which forms one part of a mmprehensive
plan for improving communications, would distribute manusaipts to special-interest
subsaiption lists shortly after submission. This has been aitic:ized as 'a vast sewer
carrying garbage from one scientist to another'. Disquiet over the very principles of the
system, which was designed to mpc with a genuine information crisis in the subject, is
mixed with mncern that the bureaucracy necessary for running it will dominate the
Association. See P. M. Boffey, 'Psychology: Apprehension over a New Communications
System', Scimet, 167 (1970), 1228-30.
23 D.]. de S. Price, Bi, Scime" Link Same" reports a paper by D. J. Urquhart on
loans of journals in the stock ofthe Science Libnry in London; about half were not used
at all, and a quarter only once in the year of survey (pP. 75-6).
SO The Varieties ofScientific Ezperitnce
or a theory now clearly recognized to be in error. But once he closes the
laboratory door behind him, he can indulge his fancy all he pleases..•. ~4
The most remarkable thing about this argument is that it assumes no
special virtue on the part of the scientist; but it does claim that the
system of quality conttol in science is perfect. The claim may have
been plausible at some time in the past, but it is no longer so.
Shoddy work exists, and in large quantity. References to it can be
gleaned from published discussions of the state of particular fields.
And it is a truly pathological symptom of the social condition of
industrialized science.
For a paper to be published, it is sufficient that the author, an
editor, and a publisher all find some purpose served by its publica-
tion. From the side of the publisher, it is a matter of economics:
given the guaranteed library subscriptions and the economics of
journal publication, it is possible to make a profit even on a obscure
journal.~5 The editor may receive an honorarium from the publisher,
and will certainly derive prestige at his own university by virtue of
his position. The author may need to have another tide in his record,
as a demonsttation (for his employer) of his continued competence
in research, or as another point to be included in his aggregate score
of publications when he applies for a grant from a large and imper-
sonal investing agency.
Of course, the publication of shoddy work would exclude a man
from membership in a community devoted to the advancement of
science; hence those who do this are either in no community at all,
or in one with different goals. In such a community, the traditional
ethic of the disciplined search for truth is either forgotten, or is a
sick joke. 26 If the publication of ~oddy science were restricted to
the dim and obscure men in remote provinces, and their local
journals, then the situation could be explained partly at least by the
natural differentiation by quality of scientists. But the participation
in this abuse by men of high prestige requires another explanation.
24 James B. Conant, 0. UtUlerstlUlllillK StinIe, (Yale University Press, New York,
1947; Mentor Boob, 1951), p. 23 of Mentor edition.
as See 'How Many More New Journals?', N""", 186 (2 Aprillg(0), 18-19-
:a6 For a penetrating analysis of the problem of shoddy science, see Howard A.
Mayerho~ 'Useless Science', BtJleti" ofllte AltnIIie Scimtists, 17 (March 1961), 92-4-
This article is a reprint of a speech originally given in 1954 by the author, a former
editor of StinIe,. He describes the various pressures on scientists to accumulate pub-
Iication-points, and gives useful case studies on how iDsignificant research can be made
fruitful and multiplied.
Social Problems ofIndustrialized Scimce 5.1
The Penetration ofScimce by Industry
It is well known that science, even the most pure science, is
important for the work of industry. Any firm engaged in modem
industrial production, and by extension an advanced industrial
nation, secures its future existence by the work done in the 'research
and development' laboratories. Such work does not merely use
scientific results; in its more sophisticated forms, it is continuous
with scientific inquiry. Problems and discoveries thrown up in such
technological inquiry can stimulate important scientific research;
and conversely, any scientific result, even one deriving from a
problem which was investigated with no thought ofapplication, may
find a use in industry. All this should be familiar; and in this
unification ofscience and technology lies the hope for the realization
of the age-old dream of material plenty for all of mankind.
What is less familiar is that industry has, in many ways, made its
own penetration into science. The industrialization of scientific
research is one manifestation of this. By itself it involves an increase
in scale, and in formal organization; but we can be sure that the vast
expense ofindustrialized science would not have been incurred by the
State unless some tangible benefit through a close association of the
two sides was expected. How close the association can be was
indicated in a report on the ethical problems of physicists in America
a few years ago."? There it was found a man could be simultaneously
filling nine roles: at his university, to be teacher, administrator, and
research scientist; with various State agencies, to be a contractor for
research, an assessor of research proposals, an official adviser on
existing projects, and a private consultant on particular technical
problems; and with commercial industry, to be a private consultant
to firms, as well as a businessman manufacturing equipment of his
own invention. It is hardly surprising that little financial abuses
should occasionally occur, nor that 'conflict of interest' should be a
matter of concern. What might be asked is whether such a man could
change his attitudes and values every time he picks up a different
piece of correspondence, or whether he operates in a world where all
distinctions are blurred, and it's all business.
The penetration of science by industry proceeds by yet another
27 See 'Hazards of Sponsored Research', Pltysics Totllly, 18 (March 1965), 98, 100;
a repon of a statement issued the previous December joindy by the American Council
of Education and the Council of the American Association of University Professors on
the prevention of conflicts of interest in govemment-sponsored research at universities.
S2 The Varieties ofScientific Experience
path, created by an inevitable ambiguity in the specification of every
research project. This is in the possible function of the result to be
achieved. Even if the scientist is personally interested only in its
significance for advancing the field, he can well imagine a possible
relevance of the result to some sophisticated and unsolved technical
problem. An emphasis on this relevance will help the project to be
considered as 'mission-oriented research', which is naturally better
endowed with funds than the useless sort. Now, it is possible for the
scientist or the investing agency to keep the different aspects of a
project, and the different values they represent, in separate compart-
ments, considering it accidental and unsought good luck that
interest in the problem should be so diversified as to bring it
financial support. But such a state of affairs cannot persist indefi-
nitely for a whole group of scientists. The natural tendency, to
ensure good relations and continued support, is to give serious
attention to those whose interest is essential and for the research to
shift in their direction.
Of course, the community of science has always had a very
difficult problem in justifying itself in terms comprehensible to the
lay world; and in its generalized claims to be the basis for progress in
industry it was to some extent playing a gentle confidence game with
its public. But it did not matter too much, for the resources devoted
to science were very small, and of them only a fraction came from
the public purse. 28 Also, the claims on behalf of the practical
importance of science were necessarily diffuse or retrospective; for
it was only rarely until the end of the nineteenth century that any-
thing like systematic scientific inquiry could be successfully applied
to the solution of industrial problems. But in the present period the
sums are significant, and the claims are true enough for those who
advance the cash to feel entitled to see some return. Hence the indi-
21 On occasion, the mnflict of values between excessively pure-minded scientists and
• particularly vulgar governmental patton muld yield disasters. See G. D. Nash, 'The
Conflict Between Pure and Applied Science in Nmeteenth-Century Public Policy: the
California State Geological Survey, 1~1874', Isis, 54 (1963),217-28. Norman Kaplan
uses the term 'bounty' to desaibe the grants formerly made by European governments
to individual scientists for the pursuit of their research; these would be on a small scale,
and recognized as conttibuting to the research work ofa university for the enhancement
of its, and the nation's prestige. See N. Kaplan, 'The Western European Scientific
Establishment in Trausition', AfI'Ieri&tm Belulvi.,,,l Seientist (December 1962), 17-21,
reprinted in Stinlee fIIUl Soeiety, ed. N. Kaplan, pp. 352-64- It should be observed that a
very important part of this 'bounty', at least between the two wars, came from American
foundations, notably the Rockefeller.
Social Problems ofIndustrialized Science S3
vidual scientist, and even more .those leaders of the scientific
community who plead for public funds for particular projects as well
as for general purposes, must be able to talk the language ofeconomic
(or military) benefit at least as well as that of the search for know-
ledge.
Influences from Runaway Technology
The relations of science with industry will not be uniformly close
in all fields; the connections will be strongest with the most modem,
rapidly developing technologies, where innovation depends entirely
on large-scale, sophisticated 'research and development'. It is these
areas, such as aerospace, eleettonics, and parts ofbiological engineer-
ing, where the pace of development is so rapid, and the ecological
and social effects so unpredictable and dangerous, that have been the
focus of public concern in the menace of 'science' to humanity.
Those who take the decisions to plunge into ever greater 'progress'
in this work are not aftlieted by any special wickedness or even
irresponsibility. They are merely continuing the attitudes and
practices inherited from the industry of the past, which might be
called 'myopic engineering'. Provided that a particular development
was technically viable and not at risk of penalties under the law, then
so long as it seemed likely to make a profit, it would be adopted with
no further thought of its consequences. Hitherto, the effects of such
a policy, however disastrous, were localized to the region where they
were put into practice; thus the rural South of England knew little,
and generally cared less, ofwhat was being done to the Midlands and
North by the Industrial Revolution. And it is undeniable that the
generally short-sighted and ruthless men who created the industry
of the nineteenth century laid the material foundations for the
prosperity of the present.
But the engine of innovation and production which we now
possess is qualitatively different in many respects. Its effects are so
pervasive that there is no place to hide from them. Also, the work of
innovation in the advanced technologies is now a large industry in
its own right. Its projects have some special features, which make
them very different from the work of inventors and scientific
consultants in the past. First, the investigation of any technical
problem ofdevelopment requires the prior commitment ofenormous
resources, both in funds and skilled manpower. Also, such problems
are necessarily speculative, in several ways. It cannot be guaranteed
S4 The Varieties ofScientific Experience
in advance that the research will produce a device which works at
all. Moreover, even if it works, there is the risk that during its years
of development, either a change in the technical or commercial
context, or the appearance of a better suited product from a com-
petitor, will deprive it of its market. Perhaps most significant of all,
the concept of 'profit' has been transformed. For much of the
sophisticated technological work is done for the State, for use outside
the market sector, as for war. Although the prospect of foreign
sales of a device are part of the calculation of 'benefit', the basic
component of benefit is assessed through a scientific study of its
potential uses. Even where a key industry is nominally in the private
sector, the State will take responsibility for its continued prosperity,
through research contracts, guaranteed purchases, and other tech-
niques. Thus a particular innovation may be recognized as risky
from the technical point of view, dubious from the commercial point
of view, of very slight use to anyone at all, even the State, and a
potentially serious nuisance to the public and source of legal and
political difficulties, and yet still receive enormous sums from the
State because of its contribution to national prestige and its impor-
tance for maintaining employment and morale in a key industry. The
Anglo-French supersonic transport is a perfect case in point of this
phenomenon. At the extreme, national prestige may become so
involved in a project that all considerations ofcost, benefit, and profit
(except, of course, to the private firms doing research and manufac-
ture) are cast to the winds, and a glamorous technical project, such
as the moon-race, absorbs resources on a gargantuan scale, all in the
name of 'science'.
Although this new industry of'R. and D.' employs many scientists
(indeed, the bulk of graduates in science and technology go there
rather than into teaching or university research), its working ethics
are descended from industry, private and state-supported, rather
than from academic science. In America, the enormous defence and
aerospace industries carry on in the time-honoured American
ttadition of 'boondoggling' on Government funds; the most effective
path to super-profits being to keep the relevant Government
agencies for cost-accounting and quality-control either remote, or
weak, or complaisant. 29
at The first extended discussion of this problem is in H. L. Nieburg, /" the Nfl""
ofSlime, (~gle Boob, Chicago, 1966). A summary appears as 'R and D in the
Contract State: Throwing away the Yardstick', BulI,ti" of lhe Atomit Sti",tists, 22
Social Problems ofIndustrialized Science SS
Thus we can speak of this new technology as 'runaway' in several
respects. In calculating cost and benefit, it ignores all those costs of
a project for which it cannot legally be called to account: in par-
ticular, the degradation of the natural and human environment.
Since the combined effect of the present and future technological
developments is likely to be catastrophic, this rush onwards can truly
be considered as out of control. And in its internal workings, the
absence of that traditional discipline, crude and frequendy distorted
but in the last resort effective, of the test of a commercial market,
makes the category of 'profit' an artificial one, to be determined by
the judgement of men in State agencies, in C<H>peration with the
promoters themselves.
It is in the borderland between science and this sort oftechnology
that we find some significant pathological phenomena. The first
occurs when a contractor (individual or institutional) develops a
really big enterprise, which is most likely to be on some mission-
oriented research in a field where money is plentiful and not too
many questions are asked. 30 There then develops a research business,
making its profit by the production of results in the fulfilment of
contracts. The director of such an establishment is then truly an
entrepreneur, who juggles with a portfolio of contracts, prospective,
existing, extendable, renewable or convertible, from various offices
in one or several agencies. The business is precarious, of course, for
his only capital is in his friendly contacts with those who decide on
the allocation of funds. In such a research factory, conditions are not
usually conducive to the slow, painstaking, and self-critical work
which is necessary for the production of really good scientific
results. Hence much, most, or even all the work can be shoddy; but
(March 1966), 20-4- A recent, more general survey of American military procurement
and spending is w. Proxmire (Senator), Report from WastelatUl (Praeger, New yom,
1970). For the repon of the Lang Committee on the English Bloodhound missiles
contract see Tite Times, 10 February 1965.
30 D. S. Greenberg has invented a prototype entrepreneurial scientist, 'Dr. Grant
Swinger'. In 'Grant Swinger: Reflections on Six Years of Progress', Scimel, 154
(16 December 1966), 1424-5, Dr. Swinger describes the exploits oCthe 'National Animal
Speech Agency', which chad its origins in the President's challenge to the nation "to
teach an animal to speak in this decade"'. Previously, Dr. Swinger had advocated a new
model ofhigh-energy particle accelerator which would extend from Palo Alto, California
to Cambridge, Massachusetts. It would be shaped so as to pick up maximum local
political support. An alternative was a vertical accelerator at the unique intersection of
four States, in the West; the design yielded. very attractive beY/dollar/vote analysis.
See '1965: Herewith, a Conversation with the Mythical Grant Swinger, Head of Break-
through Institute', Scimee, 151 (I January 1965), 29-30.
S6 The Varieties ofScientific Ezperitnce
the entrepreneur does not operate in the traditional market of
independent artisan producers who evaluate work by consensus. So
long as he can keep his contacts happy, or at least believing that they
personally have more to lose by exposing themselves through the
cancellation or non-renewal of contracts than by allowing them to
continue, his business will flourish. 31
It is in such circumstances that a man of high prestige and real
talent will produce a stream of shoddy work. Too busy to do any
thinking himself, and yet requiring a steady stream of publications
as a proof ofhis continued competence, he will toss off pieces, either
alone or with associates, which will produce a list of titles of the
necessary length. Although large-scale science is more exposed to the
risk of invasion by enttepreneurs, size is not the determining factor.
Whenever a research contractor, however modest his plant, sets the
goals of his establishment to be the renewal and extension of con-
tracts rather than the achievement of worthwhile results, he is an
entrepreneur.
Even when scientific work of good quality is being done, the style
of runaway technology can infect a field; the old, diffuse ideal of
material benefit gives way to something more sharply defined and
intoxicating: the possibility of the creation of new technical powers.
The patent dangers of some of these powers, in the present state of
civilization, have been brushed aside as of little consequence, or as
the responsibility of someone else. It is so many generations since
people in our civilization believed that there are 'secrets too powerful
to be revealed', that a scientist of our age cannot conceive himself as
being in the position of the sorcerer; and yet he is. Thus 'reckless
science', as a special product ofthe technical and social conditions of
scientific inquiry in our time, must be identified and controlled, for
the safety of humanity in the long run and for the preservation of
science in the short run. 32
31 For a satirical but none the less penetrating acmunt of the techniques of entrepre-
neurial science, see H. Miner, 'ResearchmaDship: the Feedback of Expertise', Hurna"
Or,aiutitm, 19 (1960), 1-3. The reference comes from J. Barzun, S&ien&e the Glorious
E",en';",.", (Seeker Ie Warburg, London, 1964), where this and many other targets are
put under heavy fire.
32 The leading candidate for the status of 'reckless science' at the time of writing is
mo1ecuJar genetics, which can enable the controlled manipulation of human genetic
material. See S. L. Luria, 'Modern Biology: a Terrifying Power', The Nation (20
October 1968), 406-9. One of Luria's most promising students, James Shapiro, has
taken the message and left science. He pve three reasons, of which the first is his fear
that his rauIts will be 'put to evil UICI by the men who mntrol science'. See J. K. Glass-
man, 'Harvard Genetics R.esearcher quits Science for Politics', S&ien&" 16,(1970 ),963-4-
Social Probltms ofIndustrialized Science 57
Finally, the demands of military technology in particular provide
opportunities for employment ofscientists on research projects whose
intended application lies beyond the pale of civilized practice and
morality. The weapons called 'ABC'-atomic, biological, and
chemical-are each, in their own ways, morally tainted. Research and
development of such weapons can be plausibly justified in terms of
defence and deterrence; but the .experience of the scientists on the
original atomic bomb project shows that once the weapon is available,
the tender consciences of the scientists who created it will not have
much influence on the decisions on its use.
These four abuses, shoddy science, entrepreneurial science,
reckless science, an'd dirty science, are distinct in their natures, but
there will be tendencies for them to overlap in practice. 33 Also, each
of them arises from conditions inherent to the situation of contem-
porary science; and there is no clear line of demarcation, in the
results of the work or in the attitudes of the scientist, whereby one
can condemn one man and exonerate the next. But because they
are more closely related to the demands of modern technology, and
sometimes more easily popularized as exciting than traditional
research, they will tend to attract a lion's share ofthe available funds,
thereby providing the most attractive career prospects for recruits,
and drawing into their ambit those who could not otherwise carry
on their research. This effect can be seen in the United States, where
the total budget for 'science' is enormous, but where all save a small
fraction is allocated to military R. and D. and the space-race. Even
there, it can be argued that Congress would never allocate more than
it does for the direct support of the esoteric and peculiar activity of
pure research; but the result is that the 'scientist' is seen as costing
a lot of money to the taxpayers, and if he wants to use some of it, he
is under pressure to make his accommodation with those who con-
trol these branches of runaway technology.
Information
It is impossible to separate off the different phases of the in-
vestigation of a problem into discrete and independent units. Even
in the production of data, the later stages of work on the problem
are present in an embryonic form, as expectations on the character
of the data and its refined products, and tentative plans for the later
operations. Throughout the course of the investigation of a prob-
lem, there is a continuous re-cycling, so that the problem itself
evolves, or perhaps is desttoyed, through the interaction of the
materials which it brings into being. But there are natural points
of transition, where the satisfactory accomplishment of the tasks at
one stage makes possible the invesqnent of resources in the con-
centtated, systematic work of the next stage.
The natural successor to the production of data is its refinement
into a more reliable and useful form. For the reports on the proper-
ties of particular things and events, however sound, are still far
from being 'facts'. Their ttansformation involves a new set of craft
skills, with the application of new tools, and the making of a new set
of judgements. We can illustrate this next phase of the work, the
production of 'information', with our earlier example of readings
taken off a piece of experimental apparatus. What we do may in
practice be quite simple, but in principle it is a momentous step.
Most commonly, we plot the readings on a graph, and fit a curve to
them. This step is crucial for later work, for if we are satisfied with
the fit, we thenceforth ignore the particular readings (except per-
haps to cite them as special evidence in the published argument),
and consider the curve as the report of the properties which concern
us. Thus the data has been transformed into a new sort of material,
with the aid of certain mathematical tools. This transformation in-
volves two separate judgements, which can be assisted by special
tools, but which both involve the risk of pitfalls. The first is that
the points fit sufficiently well to a pattern, and the second is that this
particular pattern, rather than any other, is the significant one. Even
the schoolboy knows ofthe first requirement; if the points represent-
ing his readings are spattered all over the graph, then he cannot
simply draw the required curve in the likeliest place and call it a
84 The A,hievement ofS,imtiji& K1I011Jledge
day. How well is 'well enough' is a matter of judgement, frequently
assisted by statistical tools, but in the last resort depending on an
assessment of the risks and costs of errors in the particular work
at hand. 9 The judgement of whether a curve is of 'the right sort' is
more crucial and more difficult. The curve which is fitted to the
points on the graph amounts to a statement of the functional
relationship between the variables of which particular values are
recorded in the experiment. Very sophisticated methods have been
developed for assisting in this judgement, but in the last resort it
depends on the craft knowledge of the scientist, to decide which
sort of functional relation is represented by the discrete set of points
obtained from his readings. lo
If the fitted curve is of the right sort, and the points cluster
around it well enough, we may say that the information is 'reliable'.
But this is not the only test it must pass. For the statements carried
Tools
In the production of experimental data, the apparatus is easily
recognized as a special sort of tool, similar in its functions to the
tools of a handicraftsman. They are not the objects of the work, but
are the means by which those objects are created and shaped. The
tools of the scientist are not all physical equip.ment; in the refine-
ment of data into information, other tools are brought into use,
most noticeably statistics and other mathematical representations.
The physical and intellectual tools of scientific work vary in their
complexity and sophistication; some require no special skill or
training for their use, while for others the scientist must have a
craftsman's knowledge of their powers, limits and possible pitfalls.
Every field has its own special tools, of which some require an ex-
tensive technical knowledge, supplemented by craft experience, for
their use. It is for this reason that a lengthy apprenticeship is neces-
sary before anyone can embark on independent work in a developed
scientific discipline. We shall see how tools condition the work which
is done in any field of science, and also enter into the social relations
between different disciplines.
We can make a rough classification of the different sorts of
materials which can be considered as tools, starting with the physical
apparatus by which data is produced. Following on this are the tools
which actually transform data into the shape appropriate for in-
formation and test it. Such data-processing tools can be statistical
techniques, or machinery which is programmed to exhibit patterns.
Science as Craftsman's Wori 89
There is a third class of tools, which do not actually produce or
transform data, but which make possible the interpretation of the
data or of the information. One example of this class is a corpus of
standard information about the objects of the inquiry, assembled
in handbook form. In sciences dealing with large collections of
different objects, noticeably the descriptive sciences, the need for
such tools is particularly strong, and there will develop entire
sub-disciplines devoted to their production. From this corpus of
standard information, materials may also be taken for inclusion as
evidence (after the appropriate tests) in the argument of the prob-
lem. In this same class are 'tool-subjects', other fields of science
which must be mastered to some extent in order that competent
work can be done in the given field. The relation of being a tool-
subject to another field, which links many pairs of fields, is in
general asymmetric; those fields which deal with the more abstract
and general properties of matter function as tool-subjects to those
less so. This asymmetry is of great significance for the development
of science over the last few centuries, and will be examined more
closely.
Finally, there is another important class of tools, of a significandy
different character. These are required when the language of the
argument of problems in a field is not merely the vernacular en-
riched by technical terms, but when it is a formal, artificial language
of its own. The most common example of this is the use of the
objects of some fields of mathematics (as the calculus) in theoretical
sciences. Such fonnallanguages are not merely a ttanslation of the
vernacular. They provide power to the arguments, making possible
inferences which could never be made in the vernacular; and also
impose a precision (not necessarily accuracy) on the objects of the
arguments. Also, the potentialities of the formal language for the
construction of arguments are revealed by continuing research in
the formal system as an object in its own right by specialists; and
the avoidance of pitfalls in its use is learned partly by craft ex-
perience but also by deeper knowledge of the mathematical system.
Hence the scientist who uses some branch of mathematics as a
language-tool must render himself competent in its practice and
theory.
Although tools are auxiliary to the advancement of scientific
knowledge, their influence on the directions of work done is im-
portant and frequendy decisive. New tools make possible the
90 The Achievement ofScientific K"lIJl)ledge
production ofentirely new sorts ofdata and information. As the tools
of a field develop, projects which previously required outstanding
talent or great perseverance, now become routine and nearly trivial.
This effect can be seen even by students, who may first study
physics without the calculus, and later discover how the laborious
and roundabout methods of describing and deriving various effects
are rendered unnecessary by this powerful tool of argument. Also,
in fields where the data comes from experience, it must be produced
in a certain general form, so that the appropriate tools can be applied
to it. This is most obvious in the case of statistical tools; many are
the cases where a consulting statistician throws back a set of data
to a surprised and somewhat hurt scientist, telling him to discard
that lot and bring back another set of a better statistical design.
In this requirement on the data we see the first of the reasons why
one cannot simply go out and get data as an independent foundation
for later work. Finally, each set of tools has its characteristic pitfalls,
and if blunders and disasters are to be avoided, the user must de-
velop a craftsman's knowledge of their properties. The abuses of
statistics are again the most notorious example of this;'. but even
in descriptive sciences, the uncritical use of handbook information,
which is always incomplete, obsolescent and not quite fitted to the
needs of the work at hand, can lead to the most astonishing blunders.
A consideration of tools also leads us directly into some important
social aspects of scientific activity. Because so many of the essential
tools for any field of science are so highly sophisticated, to achieve
complete mastery in the use of some of them involves becoming a
specialist in the tool rather than in the field to which it is being
applied. There is thus a natural division of labour between tool-
experts and their clients; and the tool-experts are not merely
individuals serving as auxiliaries to the clients in the work, but
themselves can form a self-contained speciality, a tool-providing
field. When two craftsmen with different skills are involved in the
same project, they will inevitably see the work from different
points of view. The different approaches will be complementary,
and can correct and enrich each other; but they can also be the occa-
sion of conflict. For each of the partners may be wanting to get
something slighdy but significantly different from the project. The
client wants data or information, reliable by the accepted standards
14 The classic primer in the abuse of statistics is Darrell Huff, HOJlJ to Lie Jl'it"
StlJlisti&s (Norton, New York, 1954; GoUana, 1958).
Science as Craftsman's Wori
91
of adequacy of his field, with a minimum of expense and delay;
while the tool-expert, unless he is completely subservient, will be
looking for opportunities for developing particular tools in which he
is interested The two parties even perceive the situation differendy,
for their different interests correspond to different bodies of craft
knowledge; and unless both parties enter the relation with consider-
able mutual comprehension and respect, only their respective incom-
petences will be communicated, and conflict will ensue. How such
conflicts work out depend on the social relations, and relative
strengths, of the clients and the tool-experts, individually and as
recognized collectives. It must not be thought that the tool-experts'
function as auxiliaries keeps them in an inferior position; they are
not merely essential for the work, but they may frequendy have
command oftechnicalities which are incomprehensible to the clients,
and which also bear prestige in themselves. This is most noticeable
where the tools are mathematical in character. IS
Not all tools are so specialized as to require a ~istinet corps of
experts for their use. In such cases we may speak of tool-users, not
15 The development of electtonic computers could be the subject of a useful case-
study in the history of modem technology, if attention is paid to the different purposes
of the users, the experts, and also the manufacturers. Although the development of the
operating characteristics of these machines can be seen as a steady improvement with
occasional forward leaps, their history from the users' point of view can be seen as a
series of partially resolved crises. On this, see C. Strachey, 'Systems Analysis and
Programming', Scimtiji& AmericlIn, 21S, NO.3 (September 1966), 112-26. The diverg-
ence between the purposes ofusers and experts can be seen in the criterion of 'efficiency'
of a computer program; whether it is assessed by the proportion of 'computing' time
in the total length of a program, or by the convenience of the user. Inexpert users do
not constitute a coherent group, and so their convenience does not rank high among the
specification of the functions of the computer. Only occasionally can they even publish
their criticisms, since to do so requires a level of expertise which resolves the individual
difficulty. One such publication is M. Hamilton, 'Computers for the Medical Man: a
Solution', British Metlicill JOUNUI1, 2 (30 October 1965), 1048-50. The author had found
that the so-called 'standard' programs for simple data processing are nothing of the
sort; and succeeded in devising some which could be used and interchanged without
extensive re-writing every time. The deficiencies of computers from the users' point of
view will be discussed quite frankly by manufacturers on some occasions, but these will
tend to be in the context of advertising an improvement which obviates the previous
difficulties. Thus Hewlett-Packard describes the prevailing situation asjollows: 'Given
the chance, the computer can live up to its promise. But in all too many labontories,
the computer doesn't even stand the chance of a trial because it aeates new problems
that some scientists consider to be worse than the old. ouer among these is the com-
plexity of putting the computer to work in the laboratory-programming it, mastering
the instrument-computer and the man-machine interfaces-which, to the scientist, is
often a greater drudgery than the manual data gathering and calculation that the com-
puter eliminates.' See Science, 167 (1970), 1196-7. They offer small, instrument-
oriented digital computers, and also shared time with packaged programs.
92 The Achievement ofScientific KnoJlJledge
merely clients, and of tool-using fields in relation to these products
of the tool-providing fields. There is a general asymmetry between
tool-providing and tool-using fields, which, when combined with the
prestige-ranking among sciences that we have inherited from the
past, produces a systematic tendency for tool-providing fields to
assimilate those which they nominally serve. The relation is a simple
one: those fields which describe the more abstract and general
properties of matter serve as tool-providers for those which are
concerned with the more concrete, and more particular or more
highly organized. The sequence runs from mathematics to physics,
chemistry and biology; attempts to tack the human sciences neady
on to the end of the scale have not as yet proved successful. The
reasons for this asymmetry are twofold. First, the tools which are
designed for the production of data or information in the more
particular and organized properties of matter, must necessarily be
designed to cope with complexity and variability which is en-
countered in such situations, rather than to attain the precision
which is possible in simpler and more stable contexts. Hence tools
designed for biological work will find no use in physics. On the
other hand, some of the more abstract and general properties of
matter may well be relevant to a problem in a field such as biology,
especially in the production and testing of data and information;
we have seen this in the case of statistics. This asymmetry is re-
flected in the teaching of the various sciences; the physicist's tool-
subjects are mathematics and certain techniques of instruments,
while the biologist must learn something ofall those sciences further
up the scale.
This asymmetry of provision and use does not imply that tools
are created within the confines of the more abstract disciplines, and
then hired out to those lower down on the scale. For it will fre-
quendy happen that a tool is brought into being to perform a func-
tion in a problem in a 'complex' field, is then seen to be capable of
extension in all directions, and may even gready influence the
development of the tool-providing field itself. The history of
statistics, many of whose most powerful techniques were developed
in just this way, is a case in point; and many (although not all)
branches of mathematics have evolved directly out of tools designed
for special functions in the natural sciences. 16
16 The advances in statistical "technique associated with such names as Galton and
R.. A. Fisher were achieved in the aeation of tools for analysis of numerical data in
Science as Craftsman's Work 93
We have already seen that in a general way the tools which are
available define the range of problems that can be studied. But the
influence of tools on a field can be more subde than a mere creation
of possibilities. The extensive use of a tool involves shaping the
work around its distinctive strengths and limitations; one can rarely
apply a new tool to an existing stream ofresearch without modifying
the stream strongly. Hence as new tools come into being, and are
judged appropriate and valuable by people in the field, they alter
the direction of work in the field and the conception of the field
itself. The men of an older generation who cannot master the new
tools may grumble that the field has been distorted or taken over by
outsiders. Whether they are right is a matter that only a later history
can judge. But any such judgement must be made in terms of an
implicit philosophy of science; and in the one which has been
dominant for several generations, a field becomes more genuinely
'scientific' as' it more closely resembles theoretical physics. Hence
the natural tendency of the available tools to modify a field in that
general direction are reinforced by the prestige of the limiting point
of the process. It is important to realize that this confluence of the
technical aspects of tools, and the prevailing metaphysic, is an
historical accident rather than an essential feature of natural
science. If the 'paradigm' natural science were to become a discip-
line like ecology, which uses the whole range of tool-providing
sciences but whose objects cannot be reduced to those of any of
them, then the social relations of tool-providing and tool-using
fields would be drastically altered.
The combination of these two asymmetries, one inherent in the
work, and the other socially imposed, has produced a false picture
of the relations of the different scientific disciplines. It is commonly
supposed that physics is not only stronger than biology as a field,
but also more fundamental in that its problems and objects are
independent of the latter. When two such fields are connected, as
in biophysics, the penetration is all in one direction. Yet this is
not necessarily so, for in the past it has been otherwise, on most
significant occasions. For example, few physicists will be aware that
the conception of the conservation ofenergy, which ttansfonned the
the biological and human sciences; they soon generated lively fields ofpurely theoretical
research, cast in a form comprehensible only to mathematicians. In mathematics, the
field of differential equations was origiDally a tool for rational mechanics; and Fourier's
series was invented and developed in the course of investiptiops into the theory of the
flow of heat in solid bodies.
94 The Achievement ofScientific Kno1lJ/edge
physical sciences in the middle of the nineteenth century, arose
partly out of biological problems which were investigated in a
consciously philosophical style. 17 The full history of the discovery
of the First Law of Thermodynamics shows that this surprising
origin was no 'accident'; but to explain it lies beyond the com-
petence of the traditional philosophy of science, and the physicist's
common sense of science on which it is based.
Finally, this discussion of tools shows how subde must be
the craft knowledge of a scientist who is a leader in his field. He
must not only be able to develop a craftsman's competence in the
use of particular tools, and have a sense of their powers and limita-
tions; but, if he is to do anything but follow fashions set by others,
he must assess the sorts of problems into which the use of par-
ticular tools would lead him and his colleagues and decide whether
they are appropriate for the best progress of work in his field.
Pitfalls
The craft character of scientific work is exhibited most system-
atically through the concept of 'pitfall', which I have already men-
tioned several times. IS The importance of this concept can be seen
from Francis Bacon's aphorism, 'What in observation is loose and
vague, is in information deceptive and tteacherous.' Leaving aside
his distinction ofobservation and information, we may ask why there
should be deception and treachery in the inference? This is very
strong language indeed. What does it mean? Whenever we extend
from the known to the unknown, we do so on the basis of expecta-
tions of what we will find; these are necessary to give direction to
our moves, and to supply interpretations of what we encounter.
These expectations are always incorrect in some measure, and so
we learn through the discovery of our errors. However, not all our
errors are so considerate as to announce themselves as soon as they
are made. It can happen that we follow an erroneous path of in-
17 For the importance of NlI'wpllilosoplUe in the inte1lectua1 background of many of
the 'discoverers' of amservation of energy, see T. Kuhn, 'Energy Conservation as an
Example of Simultaneous Discovery', in Criti&1Il Problems i,. ,Ite History of SeieMe,
eel. M. Clagett (University of Wisconsin Press, 1959), pp. 321-56, especially pp. 338-9.
The case of Helmholtz has been studied in great depth by Y. EIkana, 'The Emergence
of the Energy Concept' (ph.D. Thesis, Brandeis University, 1967).
18 I learned the concept of 'pitfall' as part of the aUt wisdom of mathematics from
my teacher Professor A. S. Besicovitch, F.R.S. He had the aphorism, 'In analysis, the
pit&lls are everywhere dense'; its full flavour can be appreciated only by those who
know the technical meaning of the latter tent1.
Science as Craftsman's Work 9S
vestigation for some time, investing great resources into its pursuit,
and only much later discover that we are mistaken. Then we realize
that our time and labour had been wasted, and perhaps our con-
fidence and reputation shaken as well. Bacon knew that we cannot
but try to confirm our expectations as we advance into the un-
known; and this is much easier if the things we find, as we go along,
can easily be interpreted in our favour. In this fashion, with 'loose
and vague' materials, we can deceive and bettay ourselves, to our
eventual undoing.
Thus the path of discovery is beset by concealed ttaps for the
unwary, which we can call 'pitfalls'. In some ways these are more
dangerous than the physical hazards from which they take their
name, for one learns only in retrospect that one has stumbled into
a pitfall at some earlier point of the work. One may produce data
on equipment which has a concealed systematic error; or the theory
of the equipment may have a hidden error which vitiates the in-
ferences from the readings; or the first sketch of an argument may
have an ambiguity or false deduction which undermines all the
reasonings established on its basis. At every stage of our exploration
of the unknown, we are at risk of being mistaken, and of remaining
in ignorance of our mistakes until irretrievable damage has been
done. 19
The encountering of pitfalls through the making of judgements
influenced by expectations is shown most clearly in the hazards of
19 Pitfalls in observation are most likely to occur when the 'data' are subjectively-
assessed extreme or nul points of a visual phenomenon. This was responsible for the
short-lived inquiry into 'N-Rays' early in this century; see D. J. Price, Samee si1l&e
Babylon (Yale University Press, 19S7), ch. 4- The identical pitfall vitiated the 'Allison
effect', whereby the clements 'Alabamine' and 'Virginium' were 'discovered'. See
F. G. Slack, 'The Magneto-Optical Method of Chemical Analysis', Jowul of tM
F'lInltlin Institute, 218 (1934), 44S-62. (I am indebted to Mrs. R. Countryman for this
reference.) It is interesting that the 'findings' in this second case were longer-lived;
even after the war American students of chemistry learned of these two elements by
their pattiotic namcs, rather than as Francium and Astatine.
Q!tantitative experiments can also be subject to disastrous pitfalls through the
scientist's ignorance of distorting effects on his equipment. The experiments ofD. C.
Miller with an improved interferometer, by which he detected an absolute motion of the
earth through the aether from 1925 onwards, were probably vitiated by temperature
effects, in spite of the warning on that point by Helmholtz to Michelson and Morley in
the 18805. See L. S. Swenson, Jr., 'The Michelson-Morley-Miller Experiments before
and after I 90S', JOUNUII oftM History ofAstronomy, I (1970),68. Temperature effects
also produced a pitfall in an experiment of otherwise classic simplicity, testing whether
photons lose or gain energy when moving against or with • gravitational field. In this
case, the pitfall was exposed by a Cambridge (Trinity College) undergraduate, who has
since gone on to a distinguished career in physics. See B. D. Josephson, 'Temperature-
96 The Achievement ofScientific Kno1lJledge
using research assistants, working essentially as technicians, for the
production of data. For the assistant will generally know what his
supervisor expects to find; indeed, he must have such explicit ex-
pectations if he is to make the first judgements on the soundness of
the data. But when unexpected and contrary results appear, he
must make a judgement on their significance, balancing his own
limited technical competence against the superior understanding of
his master, and perhaps being influenced by political considerations
as well. The natural course of action is to present information from
which the anomalous data have been expunged. If the supervisor is
concerned to have evidence derived from genuinely sound data, he
must go to the ttouble ofttaining his research assistants in genuinely
critical and independent thinking as well as in the craft of the
research. Sometimes this may be more trouble, and still not so
reliable, as doing the work oneself: This phenomenon indicates that
there may be an upper limit to the degree of division of labour, in
the dilution of the sophisticated craft skills, in worthwhile scientific
research. %0
the personal style will be realized through choices within the range
of possibilities defined by the whole body of methods for his prob-
lem. There is no conflict between a highly individual style in the
investigation of problems, and the production of results which meet
the socially imposed criteria of adequacy for the field. Through
personal acquaintance with a scientist's style, a colleague can recog-
nize his work; and a later historian can describe the style, and even
use it for the explanation of a unique achievement. 30 Of course, the
development ofa personal style is a matter of degree; those scientific
workers who produce bits of information on order will not generally
need, or be permitted, a sttongly marked personal style for their
tasks.
Some important features of the social aspects of scientific activity
can be illuminated by the concept of style. Since the personal style
of a matured scientist is influenced by his earlier experience, we
can speak of the transmission of style from a master to his pupils.
In this way one can construct intellectual genealogies; and to under-
stand a man's work it may be relevant to know who was his teacher's
Introduction
HAVING argued that scientific work is necessarily a craft activity,
depending on a personal knowledge of particular things, and on
subde judgements oftheir properties, I must now make an important
qualification. There are essential differences between the craft
work of scientific research and other sorts of human activities. For
the objects of scientific inquiry are of a very special sort: classes of
intellectually constructed things and events. Their difference from
the objects of handicraft production, or even of ordinary discourse
and action, gives scientific knowledge its special power, and makes
scientific inquiry a particularly complex and delicate social activity.
With this thesis I run counter to an influential tradition in the
philosophy ofscience, whose roots are in England rather than on the
Continent. It derives ultimately from Bacon, who believed that men
should, and could, turn to 'things themselves', and away from the
false theories used for describing them and explaining their actions.
A variant on this theme became popular in the nineteenth century,
through the writings of Huxley and Qifford: that science is nothing
other than 'organized common sense'.I With respect to the activity
of scientific inquiry, this belief in 'naturalness' is correct; in my
discussion of the different tasks and judgements involved in the craft
work producing data and information and using tools I did not find
it necessary to invoke any elaborate or formalized principles of
'method'. But in the extension of 'common-sense' from the work to
the objects of the inquiry, 1;he belief is seriously misleading. I shall
later discuss the very real and difficult problems of conttolling the
applications ofscientific knowledge; and these all revolve around the
• For example the classic work by W. K. cwrord, TM CtntIIIUJIJ Sense of1M uMl
SdnIe,s (1885; reprint, Dover, New York, 1955).
110 The Achievement ofScientific Knowledge
fact that this knowledge is esoteric, available only to those whose
long training in a field has made them expert.
Perhaps less crucial, but equally significant, this belief makes a
genuine history of science impossible. Until very recently at least,
the implicit assumption in the writings of the history of science
(especially at the level of schoolbooks and popularizations) was that
our sort of science is the obvious and natural way to study the world
of nature, merely the application of unfettered common sense.
Periods and cultures which failed to produce effective natural
science of our sort were condemned for an almost wilful neglect. By
the same assumption, great discoveries in science have been pres-
ented as the perception of the obvious, and the failure of a great
discoverer to get the thing quite right is the occasion for an apology
on his behalf. In this way, the history of scientific achievements
comes out as a succession of one-eyed men in the kingdom of the
blind. The question of why it required a great man to get even one
eye open, and still the question ofwhy he happened to come along at
just that time, are rendered incapable of being asked, let alone
answered. Such parodies of history are not merely of interest to
professional historians of science; since the published history of
science, especially at the popular level, is evidence for the self-
consciousness of science, it serves as a reminder of the necessity of
present discussion.
Intellectually Constructed Things and Events
I have already mentioned the classes of intellectually constructed
things and events which are the objects of scientific knowledge, in
my discussion of 'information'. Even in the production of data the
manipulations are with samples which are designed to serve as true
representatives of such classes, rather than with some 'real' objects
which can be known independently of an elaborated structure of
theory. Of course, these classes of things and events are intended to
relate, as closely as is possible, to the inaccessible reality of the ex-
ternal world. But they are different in character from that reality:
and from this difference derives the specialized nature of the craft
work of science, and the never-ending perfectibility of scientific
knowledge. 2
:I It is iDsufficiendy realized just how far removed are the conceptual objects, and the
problems, of scientific inquiry, from the world of ordinary experience. This absttaetion
can pass unnoticed because the teaching of science in the schools concentrates on the
accomplished knowledge concerning the natun1 world, with instances of where that
Scientific /nlJuiry: Problem-Solving on Artificial Objects I II
The Argument
This link between the 'material' component and the 'final cause'
of a problem is formed by the argument. Through it the experience
is made relevant to the properties of the objects of inquiry. The
argument will be a lattice-like structure of assertions about the
objects of the work, connected by inferences which are accepted for
the particular linkage functions they perform. Each assertion must
be based, either directly or indirectly through these inference-links,
on a statement ofexperience or on an explicit postulate. We shall see
later (in connection with the 'adequacy' of solutions of problems)
that the argument must have a structure that is both complex and
120 The Achievement ofScimtiji& K"ollJledge
subtle; for there is no formally valid pattern of argument that can
establish properties of general classes from reports of particular
experiences. The argument may be partly mathematical and deduc-
tive, but (outside purely theoretical fields) it must also include
inductive, confirmatory, probabilistic, or analogical inferences,
which are never capable of carrying certainty from premiss to
conclusion. l :& Because of this lack of demonstrative certainty, the
argument will include subsidiary arguments on the strength of
particular inferences, especially for those that have a crucial position
in the structure.
First, they cannot yield true and certain conclusions about the
objects of inquiry, and still less about the external reality which they
are intended to represent. This limitation is revealed by the sub-
sequent history of any solved problem, even those which seemed (in
their popularized versions at least) to establish irrefutable properties
of natural things; I shall discuss this at greater length in connection
with the evolution of scientific knowledge. Second, the necessary
complexity and subdety of a scientific argument makes impossible
any trivial testing ofits adequacy, as is possible through the examina-
tion of schematic structure in the case of simple syllogisms.
Rather, the testing is done by an application of the criteria of
adequacy accepted for the field. This itself requires judgements;
and I shall discuss it more thoroughly in connection with
'methods'.
At first glance, an argument presented in a research report may
appear to have a pyramidal structure, with a number of reports of
experience, citations of known information, and perhaps postulates,
all leading to a conclusion which consists of one or a few assertions.
But this reflects the conventional style for such reports, rather than
a real completeness of the argument. At a variety of points, the
structure of such an argument could be developed to yield further
conclusions; and one of the paths to the investigation of descendant
problems involves doing just this.
12 The complete set of such inferences is present, if implicidy, in any scientific paper
where experimental results are obtained and also explained. Inductive inferences are
involved in the taking ofthe sample studied as a ttue representative of its class; probabili-
stic inferences are involved in any assessment of the 'goodness of fit' of data; deductive
inferences are the basic links in a verbal or mathematical argument; analogical in-
ferences are used for relating the concepts understood by the mathematical symbols to
the objects ofinquiry involved in the experiment; and confirmatory inferences are made
when the evidence is cited in support of the conclusions of a theoretical argument.
Scientific Inquiry: Problem-Solving 0" Artificial Objects 121
Evidence
Only in the most purely 'descriptive' of fields can the information
have the appearance of being the direct and sole foundation of those
statements about the objects of inquiry which comprise the con-
clusion. Generally, the information appears as embedded in the
argument, providing a part of the basis for the conclusion. Hence
this information in the argument is not a statement of the properties
of the objects described in the conclusion; rather it is evidence
brought into the argument which as a whole establishes these
properties. The information which is selected from the available
stock for use as evidence must be subjected to further testing, not-
withstanding the fact that it was originally produced with this
function in mind. For each use of information as evidence is special
and demanding. Since no piece of evidence based on reports of
particular things and events can entail a positive assertion about a
general class, the weaker inferences which are made must be scru-
tinized in the light of the work they do.
The evidence must be of sufficient strength to support the load of
argument placed on it in its particular location. Moreover, the
evidence must be shown to fit the statement being supported, in
the correspondence of the objects of its reports with those of the
statement. A failure to make a proper assessment of the strength
and the fit of evidence before including it in the argument can lead
to pitfalls in the drawing of the conclusion. For these assessments,
the scientist again makes use of the criteria of adequacy accepted for
the field; and the interpretation and application of these criteria to
each particular case involves making judgements, using the craft
skill of the scientist.
The special category of 'evidence' is most easily recognized in
fields where problems involve both complex arguments and large
masses of information; and where the information itself does not
bear obvious credentials of its reliability and relevance. An extreme
example of this situation is in the law, where there is a highly
developed 'law of evidence' for the presentation and testing of in-
formation offered as evidence in court cases. In the disciplines dealing
with human history, the pitfalls which beset the inferences made
from information in its use as evidence are a recognized hazard, and
so any crucial piece of evidence must be carefully scrutinized for its
strength ilnd its fit. In most work in the natural sciences, one usually
has either a large mass of information with a relatively simple
122 The Achievement ofScientific KtJOlIJledge
argument, or a complex argument needing evidence at only a few
points. Hence neither 'descriptive' nor 'theoretical' natural sciences
generally require highly developed skills in testing evidence beyond
the tests for reliability and relevance already involved in producing
information. As in any other sort of'work, when there are fewer pit-
falls encountered and fewer special skills required the achievement
ofreliable results becomes easier; and hence more effective work can
be done by the same level of talent. Thus one reason for the greater
power of the sciences of nature, compared to most of the sciences of
man, lies in the simpler character of routine work, and the corre-
spondingly greater competence of mediocrity. 13
But even in such well-established fields pitfalls can be encountered
in the interpretation and use of evidence. For example, the con-
temporary fashion for using mathematical materials at every possible
point of an argument induces a tendency to accept statistical in-
formation as facts, rather than as evidence, in a wide variety offields.
This can lead to negligence even in testing the reliability of such
information in relation to the populations whose properties it
describes. The pitfalls in the way of even the humble tasks of
statistical data-collection are well-known to those who make their
living by the quality of the inferences based on such materials (as
commercial market surveys) but not always so well-known to others.
Moreover, the relevance of the classes established for the collection
and processing of data to those of the argument ofthe problem is by
no means automatically assured. Even when the problem is a purely
'empirical' one of determining a statistical property of a population,
13 The classic paper on the determination of the values of physical constants, R. T.
Birge, 'Probable Values of the General Physical Constants', Revietl1s ofMotlem PAys;,s,
1 (1929), 1-73, shows how subtle is the evidence, and how complex are the arguments, in
really fundamental physical research. In his introduction (pp. 1-7) he lists the common
pitfalls in the conversion of experimental data to what I call 'information': mnfusion in
the sense of 'experimental error'; the fitting of curves or lines to graphed points by
estimation nther than by calculations; and the use of incorrect values of the 'auxiliary
constants'. The argument will involve a discussion of the known sources of inaccuracy
in the data and information, as well as mmplex calculations for deriving the final results.
He stresses that the experimenter (as well as the reviewer) must exercise his judgement
at every stage. Finally, he observes that some painstaking experimental work, which had
produced accepted results, was in fact vitiated by one or another error: and, in some
cases, it was even impossible to 'save' the experimental data because the loss of the
records of its computation (including special techniques and values of auxiliary mn-
constants) prevented a ~working of it. In D. Lerner, Evillmte llrulI"fermee (The Free
Press, Glencoe, 111., 1959), there is an essay on 'Evidence and Inference in Nuclear
Research' by Martin Deutsch (pp. g6-106) which shows how indirect and mnditioned
both by theory and by expectation is the data of hig~ergy physics.
Scimtiji& /tllJuiry: ProlJlem-Solving on Artiji&;a/ Objects 123
limit (if not beyond) of the capabilities of the existing tools for
producing reliable information, the assessment of the strength and
fit of the evidence can become very subde indeed. l ? Then there can
arise disputes about the adequacy of the solution to the problem,
which cannot be resolved either by a scrutiny of the data and in-
formation, nor by an appeal to accepted criteria of adequacy. Thus
at such points, this aspect of the 'objectivity' ofscientific knowle4ge,
which is really a result of a successful social ttadition of producing
and testing the materials of that knowledge, breaks down. In the
long run, to be sure, further work will decide the issue; but the
decision on whether to engage in such further work, which partly
depends on the assessment of the adequacy of the controversial
piece, must be taken now. Thus, at such infrequent but critical
junctures in the advancement of science, the assessment of the
evidence adduced in an argument becomes a crucial judgement, in
which the individuals are thrown back on their own personal re-
sources. They are forced to put themselves at risk in making the
judgement, and they lack the safe channels of an accepted ttadition
to steer them towards the correct answer.
The Conclusion
At the end of an argument, comes the conclusion; and this is the
completion of the cycle of a research project, which is the first step
towards the achievement of scientific knowledge. The artificiality
of this product of craft work should by now be obvious. The
conclusion is not concerned with 'things themselves', but with those
intellectually constructed classes which can serve as the objects of
an argument. The contact with the external world is always un-
natural and indirect; and the reports of that contact do no more
than serve as the basis for evidence which is embedded in an argu-
ment whose pattern cannot be formally valid. Although the state-
ments in the conclusion refer to explicit objects, and are capable of
being understood by any competent practitioner, and although
every stage in the argument, including the reports of experience, can
17 Pasteur's great discovery of left and right-handed crystalline forms of tartrates was
exceptionally fortunate in its circumstances; Pasteur could convince the influential Biot
of its validity by an experiment performed in his presence. Otherwise, he might have bad
difficulties, for it is only under extremely carefu1Iy controlled conditions that the chemical
reactions needed for producing the effect can go through properly; and crystals of the
sort observed by Pasteur are very rare. See A. Ihde, Tile Dewlo",."., ofModml C1mlti1lr~
(Harper & Row, 1964), pp. 322-3.
126 The Achievement ofScimti./i& K,,01lJledge
be reproduced or tested, the work as a whole has been conditioned
by personal judgements, depending ultimately on a private, craft
knowledge.
Whether such an account ofthe completion of the investigation of
a problem seems a reasonable one will depend on the experience of
the reader. For those who imagine science as the accumulation of
hard facts and indubitable truths, a 'conclusion' as described here
will appear to be a miserably weak result, a parody of the achieve-
ment of scientific research. Indeed, it is hard to imagine how the
magnificent edifice of our established scientific knowledge could
be composed of such feeble elements. We will later see that in fact
it is not so constituted; for the conclusion of the argument, along
with the other components ofa research report, enters a new cycle of
development, in which it is further tested, and also transformed, so
that its weaknesses are exposed and if possible corrected.
But for those with either an experience of scientific work in any
lively field, or some knowledge of the history of science, this de-
scription should seem a fair one. Indeed, through an understanding
of the inherent limitations of the conclusion of a problem, we can
appreciate the inevitability of error in scientific work, at least in the
productions ofall but the men ofthe greatest genius. For a conclusion
can be no better than the evidence on which it is based, and the
objects in whose terms it is framed. The objects of enquiry delimit
the sorts of problems that can be set; and to the extent that these are
capable of enrichment and deepening, then will later problems be
correspondingly improved, and their solutions more powerful.
Similarly, the evidence depends on these, and on the tools available
at any time; and these too improve. When there is a debate over a
scientific result, neither side can draw a blank cheque on what might
be produced in its favour were better tools available; the issue must
be decided on the evidence that is there for scrutiny. A new set of
tools, providing new and better evidence, may well tip the balance,
so that while it was reasonable and correct to support one solution
under the old conditions, the new ones require a change of position.
Such changes are not easily made by a scientist in mid-career, of
course: but that is another problem, of a practical sort.
From this argument, we can derive the paradoxical conclusion,
that on occasion it has been 'Correct' for scientists to adhere to an
'incorrect' theory, even when offered what we now know to be
'correct'. For the philosophy of science of the nineteenth century,
Scientific Inquiry: Problem-Solving on Artificial 01Jjects 127
novelty are not those which existed when work on the problem began.
This is not a purely philosophical point; for in such problems, there
is a need for a frequent review, and a re-setting of subsidiary
problems, lest the work done at the earlier stages' of the project be
irrelevant to the needs of its conclusion.
Thus in a significant scientific problem, the objects of inquiry are
themselves plastic. They can remain unchanged throughout the
investigation only when the scientist is unusually prescient when
he sets the problem, or when the problem itself is banal. It is for
this reason that the investigation of real scientific problems cannot
be reduced to 'asking a question of Nature' with the expectation ofa
simple answer, or 'testing an hypothesis' to see whether it is ttue
or false. Even the description of 'normal science' as 'puzzle-solving'
canies the connotation that the puzzle is there in advance, and also
that it has a unique solution.:&1 Much of the routine work done by
scientific workers is such puzzle-solving, on essentially technical
problems. But good scientific work, even when it is not revolutionary,
is more demanding, and more interesting, than that.
The work on a particular problem is completed when an argu-
ment, meeting the accepted standards of adequacy, can be framed
and conclusions drawn. But this does not occur suddenly, as with
the dotting of the last i. Rather, the cyclic interaction of the various
materials of the problem decreases in intensity, the argument is
stabilized, the evidence becomes sufficiently strong and well-fitting,
and fewer lots of new information and data are required. The un-
expected results decrease in importance, pitfalls become negligible,
and there is a sense that the whole process is 'converging' towards
solution. But this does not always occur; and a warning sign of the
imminent failure of a problem is when the difficulties begin to
'diverge'; when the subsidiary problems called into being by new
difficulties become larger and more fundamental, and ever more
extensive modifications of the argument become necessary, with
increasing lots of fresh data being required for throwing into the
breaches. An important part ofthe craft skill ofa scientist is to sense
whether a problem is beginning to converge as early as it should, and
to detect signs of incipient divergence; and then to decide when to
abandon a doomed venture. It is the lack of such a skill that takes
21 These terms derive from T. Kuhn, Tile StnI&l"e of Scimtijie Revollllions, ch. 4t
'Normal Science as Puzzle-Solving'; the dichotomy between such 'normal' science and
'scientific revolutions' expressed in that work is, as Kuhn soon saw, somewhat too sharp.
132 The AclJieTJemmt ofScientific K1I01IJledge
beginning research students through to hopeless muddles, and on to
the final despairing struggle to salvage something of value from the
wrecbge of a failed problem.
'Stimtific Pro1J/nn' Defined
Having identified the constituents of a completed scientific
problem, by analogy with the four 'causes' of Aristotle, and also
having sketched the cycle of operations by which a problem is in-
vestigated and solved, we can now attempt a formal definition of the
term 'problem' itself. This term is becoming common, both in the
description of scientific work and in its philosophical analysis. The
concept requires some closer analysis, both for its philosophical use,
and for the distinction between 'scientific problems', which I am
discussing here, and some related, but different, sorts of problems
which I shall discuss later.
When can we say that a problem 'exists' ? We would like to be able
to distinguish between a 'problem' which is ready for investigation,
and less well-defined things, such as a sense that a certain sort of
difficulty, theoretical or practical, needs to be resolved; or that a
certain sort of work could or should be done. Any definition of the
instant of birth of a 'problem' out of a 'problem-situation' is bound
to be arbitrary, since all the components of a problem undergo con..
tinuous and possibly radical change during its investigation. Taking
our clue from Aristotle, we can say that the 'final' cause is the first
one; only when there is some specificatien of the new conclusion to
be drawn can we say that a problem exists. With this as the basis, all
the other constituents fit into place. The 'material' cause, the existing
stock of relevant scientific results as it is to be modified by the
solution of the problem, is presupposed in the statement of the
problem. Although the 'formal' cause, the argument from which the
conclusion will be derived, cannot be known in detail, the accepted
patterns of argument for the field can usually be presupposed. Of
course, in a deep investigation where this itself is modified, the
statement of the problem is correspondingly more difficult. But
something must be known of the 'efficient cause'; for a conjecture
or hypothesis which is not accompanied by a plan for its establish-
ment may be quite interesting but it is not something on which work
can be done.
Herein lies another reason why it is insufficient to characterize a
scientific problem simply as 'a question put to Nature', or as 'an
Scientific /tllJuiry: ProlJlem Solving on Artificial OlJjeGts 133
hypothesis to be tested'. Judging such questions on purely internal
features, as their surprise, improbability, or organizing and unifying
power, can lead to utterly unrealistic accounts of the evaluation of
scientific problems. 2 :& The question must contain, in addition to its
implied"answer, some plan (implicit or explicit) for the attainment
of the answer. For the solution of genuine scientific problems is not
merely having bright or even brilliant ideas; these are empty unless
they are developed and enriched by the hard, complex and sophisti-
cated craft work of scientific inquiry. Unless there is some idea of
how the work will be done, there is no way of knowing whether the
solution can even be achieved; and in general the form that the
tentative solution takes will depend on the projected means of its
accomplishment. This is of course a matter of degree; as a limiting
case there is the handful ofclassic unsolved problems in mathematics.
But these are accepted as genuine problems either through the
naturalness of their objects, the empirical evidence for their truth,
or through the authority of their proposers.
The use ofthe term 'problem' to define the programme ofinquiry
goes back to antiquity in mathematics; although even there it has the
connotation of being able to do something rather than to know some-
thing. It has occasionally been used to give point to a precept of
method in other fields, the most famous example being Lord Acton's
dictum, 'study problems, not periods'. But it has tended to carry
with it the implication that solving a problem is a form of inquiry
which is both safer and less deep than something else which is being
abjured. It would be most unfortunate if this present discussion
were taken as evidence for the proposition that 'all the scientist does
is to solve problems' as if this were equivalent to doing crossword
22 Lest the very idea of'hypothesis' be lost when the so-called 'hypothetico-deductive
method' loses favour among philosophers of science, I should remark here that the
pattern of argument in statistical inference necessarily involves the report of the testing
or a prior hypothesis; and many pitfalls of interpretation can be encountered unless the
work actually proceeds by first framing the hypothesis and then getting the relevant data.
But a worthwhile hypothesis for statistical testing can only be set when the problem has
gone through its earlier stages of development, and is ready to be put in a final, crysta1-
lized form. The deep and complex problems of the validity of such real 'hypothetico-
deductive' arguments are little studied by the philosophers who have concerned them-
selves with this 'method' in the abstract. It is Doteworthy that the BritisA Jowul ftW tile
PllilosopAy ofStimtt, a leading journal in its field, has had just a single paper on the logic
ofstatistical inference in the twenty years of its publication. The exceptional paper which
proves my rule is I. Hacking, 'On the Foundation of Statistics', BritisA Jowul ftW tile
PhilosopAy of Stimtt, IS (1964), 1-27. See 'Index 19Scrlg6g Volumes 1-20', BritiiJI
Jowglfor 1M P/JilosopAy of Samet, 21 (1970), 21.
134 The AcAievemmt ofScientific Kno1lJledge
puzzles. For I am arguing that whatever a scientist does, it is best
conceived as the investigation (including both the creation and the
solution) of problems. We shall see that problems can vary in depth
from the trivial to the profound, and that when genuine scientific
knowledge comes to be, it is achieved through a complex social
endeavour, where the materiaIs embodied in the solution of one
problem are tested and transformed through their use in the in-
vestigation of subsequent problems.
We may now put our definition formally, and say that a scientific
problem is a statement (always partial and subject to evolution) of
new properties of the objects of inquiry, to be established as a con-
clusion to an adequate argument, in accordance with a plan (specified
to an appropriate degree) for its achievement. The naturalness of
this definition, for the social activity of science in the later twentieth
century, can be seen from its implicit use when applications are made
for research grants and contracts. The two components, statement
and plan, which we might call the 'final' and the 'efficient' causes of
the problem, are each necessary, and are joindy sufficient, for
specifying the project. On their basis, the judgements of value,
feasibility and cost can be made; and it is in terms of such judge-
ments that the decision is made for investment in the work. The
evolution ofthe problem in the course ofinquiry is a common occur-
rence; those who control investment in scientific work are aware
of this tendency to change, and they can also tell when one prob-
lem has turned into another.
With this definition of 'scientific problem', my analysis of the
cycle of the first, individual, phase of the achievement of scientific
knowledge is nearly complete; in a subsequent chapter I will discuss
the second, social, phase, in which the solved problem is accepted
as a research report and is then subjected to further wotk before
becoming a 'fact' and ultimately 'knowledge'. In this chapter we
have seen how the problem is solved by a conclusion being drawn
about the classes of intellectually constructed things and events
which are the objects of the investigation. The conclusion is the
outcome ofan argument, and is related to the external world through
the evidence embedded in the argument. This evidence derives
ultimately from the data, which is the sole point of contact with the
external world, through the information which is produced from it
through the application of tools. Throughout, the work involves the
use of craft skills, and the making of a variety of judgements; the
Scientific Inquiry: Problem-Solving on Artificial Objects 135
character of both of these is conditioned by the artificiality of the
objects of the inquiry.
Judgements of Adequacy
Each component of the argument of a solved problem, either an
inference-link or a piece of evidence, can be no more than adequate
to its function in the total sttucture. And what is 'adequate' will
depend not merely on its context in the problem, but on the general
criteria of adequacy for the class of such problems imposed by the
community. In the beginning of this part I gave an example of the
Methods I S3
necessity of judgements of adequacy, in the discussions of the
'soundness' of data, and of the 'reliability' and 'relevance' of in-
formation. An appreciation of what is involved in such judgements
may be gained from a consideration of a common and routine pro-
cedure in the formation of such judgements: statistical significance
tests. For statisticians do not simply say that a correlation is
'significant' or 'not significant'; rather, they will speak of sig-
nificance at a certain level. Those who have any craft skill in the
use of such tools will appreciate that the significance level to be
adopted is not assigned by God, but must be decided by the user.
The decision will be based on estimates of the direct costs and the
risks associated with each level. For each level of significance in-
volves the possibility of two sorts of error: of rejecting worthwhile
information, or of allowing dubious information to pass. The more
stringent the test chosen, the safer; but also the more costly, be-
cause of the extra time, care, and resources required for producing
material that will pass it. The choice of a particular level of sig-
nificance must depend on a judgement of what degree of safety is
required, for that component in its context in the total problem.
And this judgement must be based on general criteria of adequacy
applied to that particular situation. There can be no perfectly safe
test of the quality of the material, and neither can there be a cer-
tainly correct decision on the degree of stringency of the test.
In general, we can say that imposed criteria of adequacy are
necessary for scientific work because of the inconclusiveness of the
arguments used in science. This inconclusiveness follows from the
peculiar character of the objects of scientific inquiry: classes of
intellectually constructed things and events, the evidence for whose
properties is derived from particular experiences. To move from
the reported properties of particular samples, to the properties of
the classes they are intended to represent, a demonstration is
necessary; but a formally valid argument, yielding certainty or
truth, is impossible.
A scientific problem is thus incapable of having a solution which
is 'true'. Rather, the solution will be assessed for adequacy; and
for this every component must be so assessed. We can distinguish
two sorts of criteria and judgements of adequacy: those relating to
the argument, and those relating to the evidence. In the former
class are the tests of the various inferenceS which carry the argu-
ment through the statements of the properties of its objects to the
154 The Achievement ofScientific KnoT1Jledge
conclusions. Even a deductive argument is not exempt from testing
for adequacy; as soon as one passes beyond the simple syllogism in
sophistication, the obvious and intuitive tests fail to apply. Argu-
ments cast in mathematical form must satisfy criteria of 'rigour' to
the appropriate degree; and the rigour of physicists' mathematics
is not, and need not be, in the same class as that of the pure mathe-
maticians. Even in pure mathematics, disagreement over criteria of
adequacy have erupted at crucial points in the development of the
subject; the most.recent important case occurred at the beginning
of the present century, in connection with the use ofactually infinite
sets and constructions in mathematical proofs. 6 The criteria of
adequacy relating to evidence are more varied; for they control not
only the conditions of the production of data and information but
also the strength and fit of the evidence in its particular context. It
will frequendy be necessary for some of the evidence to be explained
6 At the end of the nineteenth century it could appear to mathematicians that they
had reached evolution's end, in the perfection of aiteria of adequacy. Thus Henri
Poin~: 'Today there remains in analysis only integers and finite or infinite systems
of integers, inter-related by a net of relations of equality or inequality. Mathematics
bas been aritbmetised.••• Have we at least attained absolute rigour? At each stage of
the evolution, our Cathers believed that they too had attained it. If they deceived them-
selves, do not we deceive ourselves as they did? ... We may say today that absolute
rigour bas been attained.' (Qpoted from E. T. Bell, Tile DevelojmInJI of M.tltnutits
(McGraw-HiD, New York and London, I94S), p. 34S.) Poincare was here reflecting on
the achievements of Weierstrass and his school in establishing the very tight patterns
of argument for 'potential-infinite' proofs in analysis. But the fa~de of perfection was
already revealing cracb in two places. First, one of the most basic objects of inquiry,
'real number, had been under investigation for some decades, and this 'foundations'
work was enQ)UJltering anomalies and oandoxes at an increasing rate. These were
related to the basic structure of logic, and to the 'actual infinite', which, after its com-
plete banishment in the middle of the century, was aeeping back into mathematical
arguments. At the tum of the century there was a 'foundations aisis', which led to the
inspired works of Russell, Hilbert, and Brouwer. The sequel to this was the aeation of
a new field of mathematics, in which many problems were set and solved, some of them
very deep (G6del's theorem being the outstanding example). But the basic problems
are still far from solution; see I. Lakatos, 'Infinite Regress and the Foundations of
Mathematics', AristotelitM Soeiet7 S.,plmtetJtMy VollIIM, 36 (1g62), ISS-8.f.. For a
personal acmunt of the reaction to the failure of all of the programmes to establish
'foundations', see John von Neumann, 'The Mathematician', Colletted Woris, vol. i
(Pergamon Press, 1961). He remarb that after the failure had been perceived by working
mathematicians, they 'decided to use the system anyway', for 'it stood on at least as
sound a foundation as, for example, the existence of the electron' (pp. S-6). Thus
working mathematicians operate with respect to the 'foundations' of their subject much
IS engineers do with nineteenth-century rigorous analysis: their techniques are designed
to avoid the known pitfalls, and no new ones have been detected. Since it is indisputable
that the 'foundations' of mathematics are not established, the absence of any new
'foundations crises' in the past haIf-<entury might be considered as evidence of a lack
of deep amc:eptuaI development in that period.
Methods ISS
and defended explicidy; and these subsidiary arguments must also
meet criteria of adequacy appropriate to their function. Thus the
complexity of a solved problem is matched by that of the set of
relevant criteria of adequacy; and that set will depend closely on
the field of inquiry. Hence it is impossible to produce an explicit
list of criteria of adequacy applying to a wide class of problems.
This component of 'method' is thus a craft knowledge, but no less
essential to successful scientific inquiry than the formal, explicit
knowledge deposited in the public record.
This brings us back to the paradox of adequacy with which this
discussion began. It would appear that the certainty and objectivity
ofscientific knowledge are dissolving in a set of intuitive judgements
based on. principles which frequendy cannot even be made explicit,
let alone defended or tested. The easiest escape from the paradox
is to show that it is a false one; that the situation is not at all one of
objective truth being based on subjective guesses. For we have the
historical knowledge that some fields ofscience do achieve objectivity
and near-certairtty in their results, while others do not. In all of
them, there is an absence of formally valid proofs, and a presence
of conttolling criteria of adequacy. The difference between them
does not lie in this logical aspect of their arguments and conclusions,
but in the particular circumstances of their development.
The criteria of adequacy associated with a field are, however,
directly relevant to the sttength it attains. Indeed, they perform
an essential function, and a field is matured and effective only when
they perform it well. There the deficiencies created by the lack of
formally valid patterns of argument for basing general conclusions
on particular reports are remedied for practical purposes by the
accepted criteria of adequacy. For the tests of adequacy imposed on
the components of a problem at each phase of its development serve
for the avoidance or the timely recognition of the known pitfalls
in the way. In a matured field, unknown pitfalls are relatively rare,
and so the conclusions of the arguments will be well-established.
Although they are still not immune from the eventual discovery of
error or inadequacy, they provide a firm basis for immediately
subsequent work in the field. Also, the social tests on the materials
of a research report, necessary for their becoming accepted
as 'fact', can be applied in a sttaightforward manner, and a
direct transition from one status to the other can be the normal
event.
156 The Achievement ofScientific Knowledge
C;;terUl ofAtlefJUII&Y and the Maturing ofa Discipline
The function of criteria of adequacy is thus to make 'facts' p0s-
sible. It is worthwhile to examine this more closely; we can do so
in terms of the maturity or otherwise of a field of inquiry. The
condition of ineffectiveness or severe immaturity in a field can be
recognized (by those who wish to do so) through several symptoms.
In such a field, controversies on results range indiscriminately and
inconclusively from criticism of raw data to abstract methodology.
(Whether such controversies are common, depends on the social
character of the field; roughly, whether it is nascent or moribund.)
In this we see an effect of the weakness of the accepted criteria of
adequacy; they are not strong enough to channel debates on to
well-defined problems, nor to provide an agreed foundation for the
debates. Another symptom is that 'facts' do not exist. There is no
cumulative development of accepted results, stable under testing
and repetition, and invariant under changes in the problems in
which they are used. Instead, the results of each school or tendency
die with the problem around which its work was organized.
These two symptoms are related through a third, the ubiquity of
pitfalls. The failure of facts to be achieved is an effect of undetected
pitfalls. The conclusions of nearly all problems are soon seen to
contain errors; and although this is no hindrance at all to further
research in the same line, their materials are then valueless as in-
formation to be used in other work. But to conttovert a result is
not the same as to identify that flawed component in the argument
which vitiated it. Such flaws are rarely errors in logic; for as we
have seen, the basic patterns of argument in science are not formally
valid logical ones. The flaws are truly pitfalls, false assessments of
the quality of evidence or insufficiendy strong inferences; and they
are not easily identifiable even in retrospect, for the pattern of any
argument in a real scientific problem is too complex to permit of
an easy dissection. The presence of such pitfalls is caused by in-
sufficiendy strong criteria of adequacy. For when a particular pitfall
is identified, it is signposted by a particular criterion of adequacy:
at this point the work must be done rigorously, in a particular way,
lest disaster result. Conversely, a weak set of criteria of adequacy
fails to guard against pitfalls, and yields a situation where conclusions
of problems are illusory, and facts cannot exist.
The judgements of adequacy thus perform the same function in
scientific inquiry as the tests for quality control in industrial manu-
Methods 157
facture; and the criteria ofadequacy stand in relation to them as the
standards of acceptable quality set for the tests. The solution of at
scientific problem is like a complex and delicate manufactured
product, in that a single weak component can destroy the whole
work. In manufacture, as in science, perfection is impossible; and
the standards of quality must be determined in the light of the
experience of the costs and risks associated with each level. The
difference between manufacture and scientific inquiry, in this con-
text, lies in the uniqueness of each scientific problem; even its
components are not all uniform, easily tested objects. Hence the
standards of quality control for scientific work are not capable of
being set out in handbook form, nor the tests carried out by a rou-
tine. The criteria and judgements of adequacy are necessarily an
informal, largely tacit knowledge; and in mastering and applying
them the successful scientist is a sophisticated and highly skilled
craftsman.
In the work of bringing a field towards maturity, an important
part lies in the strengthening of the criteria of adequacy. This is
not all, of course; the development of new tools, and the creation
of an appropriate social environment, are equally important. Nor
can the strengthening of the criteria of adequacy be done in an
abstract, automatic fashion, as by the attempted imitation of a
successful field. Such a strategy can produce totally misdirected
criteria of adequacy, leaving a field in worse condition than it was
originally. For the relevant criteria of adequacy are, as we have seen,
intimately related to the characteristic pitfalls of the problems in-
vestigated in the field; and these will be very particular to it, in its
objects of inquiry, its sources of data, its tools, and its patterns of
argument. Moreover, each particular criterion of adequacy must,
if it is to be effective, carry with it the craft knowledge of making
judgements of whether it is satisfied in particular cases, as well as
the tools and techniques for preparing materials of a quality
sufficient to satisfy it.
Over recent centuries, the founders and leaders of ineffective
disciplines have believed and proclaimed that the application of the
correct 'method' would quickly and automatically bring their
studies to a state of maturity and effectiveness. The conception of
method which has generally prevailed has been a combination of
two elements: a simplified pattern of argument (either inductive or
hypothetico-deduetive, depending on the current fashion), and a
IS8 The Achievement ofScientific Knowledge
commitment to simple, preferably quantifiable data. Leaders of
influential schools in the social sciences have sincerely believed that
real science is done by putting masses of quantitative data through
a statistical sausage-machine, and then observing the Laws which
emerge.7 From such caricatures of the process of scientific inquiry
are derived criteria of adequacy which enforce an apparently
rigorous procedure of research,8 but whose results can but rarely
escape vacuity. Such programmes of reduction and mathematiza-
tion base their claims on the undoubted successes of the physical
sciences since the seventeenth century; but they ignore the long
series of dismal failures in applying this approach to the sciences of
life, thought, and society. And the latter history, largely unwritten,
is precisely as old as the success story: 'l'homme machine' of
Descartes's chimerical physiology is an exact contemporary with
Galileo's successful mechanics. The principle that each field of
inquiry has a degree of precision of argument appropriate to its
objects was laid down by Aristode; and the great mathematiCian
Gauss observed that 'lack of mathematical culture is revealed
nowhere so conspicuously, as in meaningless precision in numerical
computations'.9 If we extend these principles to criteria of adequacy
, Thus, T. S. Kuhn, 'The Function of Measurement in Modem Physical Science',
Isis, 5z (1961), 161-93, starts: 'At the University of CUcago, the fl9lde of the Social
Science Research Building bean Lord Kelvin's famous dictum: "H you cannot measure,
your knowledge is meagre and unsatisfactory.'" Gnnting the importance of measure-
ment in physical science, he asserts, 'I feel equally convinced that our most prevalent
notions both about the functions of measurement and about the source of its special
efficacy are derived largely from myth' (p. 161). He argues this thesis at length; and
the reader will see the influence of this essay on my own ideas. It is reprinted in H.
Woolf (ed.), /lflMlifielllitm: • History of 1M Me"';", of MetlswtmnJI i" 1M NtllUf'tll
.tIIl S«W Snmees (Dobbs Merrill, Indianapolis and New York, 1961).
• Much of the classic study, O. Morgenstern, 0" 1M Aeewtuy ofEetnUJmie Observll-
l;tmS (2nd edn., Princeton University Press, 1963), can be interpreted as an analysis of
misdirected aiteria of adequacy. On the basis of data that is deeply flawed by inherent
inaccuracies, 'applied' economists aggregate naticmal statistics to a high degree of
pseudo-prec:ision, and 'theoretical' emnomists construct models in which these very
dubious numbers are manipuJated in sophisticated and sensitive mathematical pro-
cedures.
9 Qpoted from O. Morgenstern, 0" 1M Ae"''''7 of EetnUJmie ObSerDillions, p. 99.
The German original reads: 'Oer Mangel an mathematischer Bildung gibt sich durch
nithts so auffallend zu erkennen, wie durch masslose Scharfe im Zahlenrechnen.' My
translation differs slighdy from that given in the text. For Aristode, see Nie0ttUl&1Ie""
Elllies, tt. W. D. Ross (Oxford, 1915), th. 3, l094b IZ-28: 'Our discussion will be
adequate if it bas as much clearness as the subject-matter admits of, for precision is not
to be sought for alike in all discussion, any more than in all the products ofthe aafts....
In the same spirit, therefore, should each type of statement be received; for it is the
mark oC an educated man to look Cor precision in each class of thinp just so far as the
159Methods
in general, we can appreciate the fallacy of naive imitation as the
road to success.
One sign of a field having achieved maturity is a certain under-
lying stability, which persists through all the rapid changes in
results, problems, and even objects of inquiry. This stability is
revealed by the absence of new pitfalls, except at the outermost
frontiers of research; and, related to this, a set of appropriate and
stable criteria of adequacy. When these have remained in the inter-
personal channel of communication for some time, they become part
of the basic unselfconscious craft knowledge of the field; and the
particular judgements of adequacy which are made do not seem to
depend on anything but common sense. In these conditions the
very existence of criteria of adequacy can be overlooked; and philo-
sophers of science, basing their analyses on the experience of just
such fields, can remain in ignorance of their existence and im-
portance. But, as we have seen, in less matured fields they cannot be
taken for granted; and they will be of central importance in several
of our later discussions of the activity of science.
Criteria of Value
The criteria of value, and the judgements based upon them, form
an interesting contrast to those of adequacy. In those we saw a close
relation with philosophical questions of the possibility and nature
of scientific knowledge; while here we shall find ourselves involved
in problems of the social activity of science. The difference arises
from the functions of the judgements: the one is basically that of
assessment of work done, either a whole problem or a component;
and the other is an important determinant of the direction of future
work. For although judgements ofvalue are also made on completed
work, their crucial role is in the choice ofproblems to be investigated.
Thus, while the strength of the existing achievements of a field
depends (as we have seen) on its criteria of adequacy, its health and
future prospects are intimately related to its criteria of value. Be-
cause the criteria of value are complex, and necessarily involve
predictions ofthe future, the achievement ofappropriate judgements
of value is a delicate social task, even in matured fields where the
criteria of adequacy are well-established. In this respect, the social
problems of scientific inquiry in a field require attention even when
nature of the subject admits; it is evidendy equally foolish to accept probable reasoDiDs
from a mathematician as to demand from a rhetorician scientific proof's.t
160 The Achievement ofScientific Knowledge
the philosophical problems of scientific knowledge have been given
a practical resolution.
The exclusion of problems of value from the traditional philo-
sophy of science has its roots in the ideology of modern natural
science as it was formed through many generations of sttuggle. The
earliest conflict was on the philosophical plane; the pioneers of the
new philosophy of the seventeenth century conceived a world of
nature devoid of human properties, and accordingly rejected all
explanations in terms of 'final causes', or purposes of things and
events. Although they usually invoked the Deity in one way or
another, to give meaning to the world of nature in general, their
explanations of particular phenomena were cast in terms of purely
efficient causes. As the new science became established, the sttuggle
for the autonomy of its collective goals required an ideology which
denied any direct relation between the purposes of the community
at large, and the functions of the products of its work. Spokesmen
of science would remind lay audiences of the benefits which
ultimately flowed from their work, but would warn them that the
work must be pursued strictly for its own sake, lest the free play of
creativity be stifled. Thus science, conceived as a body of factual
knowledge, was cut off from considerations of value in two ways:
its assertions were purely descriptive, with no normative element;
and the considerations of social value by which all other human
activities are assessed were declared irrelevant. 1o The deeper
problems in which values enter, such as those concerning ethics
and morality, could for a long time be neglected because of the
favourable social conditions inside science and its very slight
responsibility for the social effects of industrial and military pro-
duction.
Certainly the facts presented in a textbook of mathematical or
physical science bear no signs ofbeing conditioned by considerations
of value of any sort. Even here, the appearance is illusory; for the
selection and presentation ofthese facts has been strongly influenced
by judgements of value. When one passes to books in the biological
sciences or applied sciences, considerations of value enter through
10 The defensive position of those who wanted to argue for the presence of values in
human knowledge is indicated by the tide of a book by the great psychologist W.
Kahler, Tile PIMt of VIII. in " Worltl of FMts (Liveright, New York; Kepn Paul,
London, 1939). A review of the litenture on this problem will be found in J. Leach,
'Explanation and Value Neutra1ity', BritisIJ]tnJnuIlfor PllilosopAy ofSrimct, 19 (1g68),
93-108.
Methods 161
Components ofValue
The criteria of value applied in judgements on scientific problems
and results do not relate directly to the Good; but they are not
thereby deprived of interest. 12 We can distinguish three independ-
ent components in these criteria, two relating to the functions of the
completed piece of work, and the other to the purposes of the
scientist in undertaking it. Of these 'objective' components, the
first is the 'internal' one: to what extent the solved problem will,
or does, advance knowledge of the objects of inquiry of the field,
either directly or through suggested descendant-problems; this is
the dominant criterion of value in 'pure science'; and we note that
it is the least amenable to precise estimation, to say nothing of
12 The classic paper on aiteria of value is that of Alvin M. Weinberg, 'Criteria of
Scientific Cloice', M;"""., I (1963), 159-71, and ibid., Z (1964), 3-14; reprinted in
Rt/leaitnu till Big S~t (Perpmon Press, 196'7), pp. 65-122.
Methods 163
reduction to quantitative measurement. The criteria of 'internal'
value adopted for a field will depend both on the social experience
of what has been successful in the past, and also on the general
conception of where the field is, or should be, heading at the point
of its development. For example, there may be a 'grand problem',
inherited from the founders of the field, whose eventual solution
defines the ruling criteria of value. 13 Alternatively, if the field is in
a phase of consolidation after a significant breakthrough, then
systematic studies of masses of particular details of a certain sort
will have considerable value, whereas previously the objects of
inquiry would have been changing too rapidly for such work to be
worthwhile. Within such general considerations of strategy, the
tactics of scientific progress are so varied as to offer a great many
possible functions for the results of solved problems, correspond-
ing to the variety of problem-situations out of which problems are
born. A knowledge of these functions, and an assignment of value
to each of them, is part of the craft knowledge of each field; and it
would be pointless to attempt an exhaustive formal list of them
covering all of science. Any attempt to make one sort of function,
such as 'explanation', 'prediction', or 'unification under a general
law', to be the unique and defining basis for criteria of value, is
bound to lead to an unrealistic account of scientific inquiry.
Parallel to this is the 'external' component of value: the con-
tribution that the completed project makes to the solution of prob-
lems, or the accomplishment of tasks, outside the given field. There
is a considerable variety in these external domains: they may in-
clude other fields of science, or technology, or (in earlier ages)
philosophy, or even the social or ideological needs ofthe community
of science or its larger society. Hitherto I have implicitly restricted
my discussion to problems whose dominant value is 'internal'. I
century using precision optical iDstruments for angle measurements, and the abstract
mathematics of group theory for theories of internal structure, was invaded in 1912 by
X-ray tedmiques invented by von Laue: and the field of X-ray crystallography soon
grew to become of the greatest significance in physics and biology. Materials for an
history ofthis field have been assembled in an exceptionally fine collection, Fift1 Yeus
ofx...,., Dijfrlltl;(JfJ, ed. P. P. Ewald (Oosthoek, Utteeht, 1962). In this, an essay by J.
D. H. Donnay, 'For Auld Lang Syne' (pp. s64-g) quotes a crystal10grapher of the old
school, A. F. Rogers, 'Geometrical crystallography has bad a glorious past, and it will
have a glorious future. t (I am indebted to Mrs. Rachel Countryman for this reference.)
Such invasioDs by tools are not always beneficial; particularly when an immature field
is invaded by quite inappropriate tools, its work can be seriously disturbed. We sball
discuss this in more detail below.
Facts and their Evolution 197
the tool itself and the development of standardized versions of it. I Z
The two processes are not exclusive; for whenever a particular
physical tool is manufactured in quantity, or an intellectual tool
described in a book, it is necessarily standardized to perform a
variety of functions, operating on different objects in various fields.
We can see the extreme case of standardization in the development
of versions of tools suitable for use in teaching. Having learned
something of the craft of manipulating with this class of tools, from
experience with his 'fool-proof' models, the student will in later
years be better equipped to cope with more sophisticated versions in
the course of his scientific or technical work.
One feature of the standardization of tools is relevant to the
general problem of this chapter. This is, that a thoroughly standard-
ized version of a tool will be longer-lived than a sophisticated,
specialized version. This longevity arises from several causes. A tool
which is designed for the most effective production of very speci-
alized data or information will be more sensitive to changes in its
function; while one which performs reasonably well in the pro-
duction of non-critical material in one problem will be likely to do
so for its descendants. Also, the standardized tool, being, designed
for functioning in a wide variety of fields, will be less affected in its
overall use by changes in anyone of them. Finally, unless fashion
and prestige-replacement are dominant in a particular scientific
community, such tried and true tools, whose possibilities and pitfalls
are well known to the amateur craftsman users, will be replaced only
reluctantly.13 The extreme case of such permanence will be seen in
12 The microscope provides a good example of a tool whose use has been extended
over all fields of science and technology, and which has developed new specialized
versions and their standardized descendants. Indeed, so pervasive is such a tool, that the
writing of its history would be an enormous task, involving the close attention to experi-
mental technique in a great variety ofcontexts. The standard recent history, s. Bndbury,
The Evolution of the Microscope (Oxford University Press, 196'7), necessarily confines
itself to advances in design in a more or less linear sequence.
A picturesque but also penetrating desaiption of the evolution of tools was given by
]. aerk Maxwell, in a review of papers by Lord Kelvin in 1872: (The reader) 'may also
study, in the recorded history of electrometers, the principles of natural selection, the
conditions of the permanence of species, the retention of rudimentary organs in manu-
factured articles, and the tendency to reversion to older types in the absence of scientific
conttol.' See Tile Scimtiji& P.pers ofJames Cieri MU7I1eli (New York, 1890; reprint
Dover Publications, New York, no date), ii, 3°4-
13 A most interesting study on the cultural influences on the exteDsion of a mathe-
matical tool is A. N. B. Garvan, 'Slide Rule and Sector: a Study in Sc:ience, Technology
and Society', Proceedings of the Tmth l",eruhtmlll Convess of Kuror, of Sdnlee,
(Hermann, Paris, 196.f.), i, 397-400. The historical problem is why the slide rule,
198 The Achievement ofScientific KnollJledge
schoolteaching, where the combination of poverty and inertia can
occasionally cause the retention of experimental apparatus for
generations after it has ceased to have any sci~ntific significance
whatever.
We shall soon return to a discussion of standardization in a wider
context; but there are some features worthy of notice in all these
paths of evolution of tools. Much useful work in science is done in
the development of tools, and even in the exploratory work ofseeing
whether they can be adapted to perform new functions. Such work
rarely involves the great conceptual advances that are the most
dramatic part of the growth of scientific knowledge. This is not to
deny that such work can require great skill, talent, or even courage.
But the production or design of things for a preassigned function,
whether they be tools or particular items ofinformation, is a problem
more of a 'technical' character than scientific; and it generally calls
for a different style of work. It is on such tasks that uninspired
'scientific manpower' can be usefully employed.
But it must not be thought that all work on tools is for the unsung
and anonymous workers, no more than that tool-experts are always
in a servant-relation to tool-users. It is not merely that the first
devising and development of very powerful tools requires great
talent and commitment; but the further development of such tools
can give rise to coherent, self-contained scientific disciplines, whose
objects of investigation are the descendants of the original tools. In
such fields, the possible functions of the objects may eventually be
reduced to a minor or even negligible part ofthe inquiry, and purely
scientific problems become dominant. The tendency for tools to
become the ancestors of scientific fields is most noticeable in mathe-
matics; from geometry itself, to differential equations and statistics,
the process has repeated itselfmany times. An analogous process can
be seen in the development of chemistry, which emerged from the
status ofa tool-providing art for metallurgy and medicine to become
an independent scientific discipline. Such an evolution of a field
entails drastic changes in its methods, and in particular in the con-
trolling judgements of adequacy and value; and with them the
invented in the early seventeenth century, remained in obscurity for 200 years before
replacing the sector or 'mathematical compasses' IS the standard hand~culating
instrument. There are many facton involved, including the permanence ofcraft tradition,
the teaching of mathematics through geometry, and the retention of a 'classical' theory
of architectural design, all until the later nineteenth century.
Facts and their Evolution 199
corresponding transformations in the character of the community
involved.
17 When facts are retailed in an older, incorrect version, it is necessary that the students
be steered away from pitfalls which were later discovered in that version; and so the
facts as purveyed become a vulgarized version of something that never existed. For
example, a consideration of 'latent heat' in the theoretical context of the eighteenth-
century work could lead an unwary student to repeat an error common then. Q!wltity
of heat was measured by the temperature change resultant on (ts inttoduetion to a body
ofgiven mass and heat capacity; and this measure was naturally extended to the measure-
ment of latent heat. At that time, there was no reason to suppose that an experimental
result as 'I have, in the same manner, put a lump of ice into an equal quantity of water,
heated to the temperature 176, and the result wast that the fluid was no hotter than water
ready to freeze' could not be ttansformed into • general measure of heat of fusion, IS
'But this quantity ofheat (which disappeared in melting the ice) would be inaeasingt by
143 degrees, the heat of a quantity of water, equal in weight to the ice alone'; and then
applied to 11"1 substance, as ' ••. the heat which any given quantity of water loses upon
being frozen-were it to be communicated to an equal weight of gold, at the tempera-
ture offreezing, the gold, instead ofbeing heated 162 degrees, would be heated 140 x 20=
2800 degrees or, would be raised to a bright retl1Jellt.' The first two quotations are from
Joseph Black, Lea"es on the Elements ofClJemistr] (Edinburgh, 1803) i, 124-5; the third
is from Rumford, 'An Inquiry concerning the weight ascribed to Heat', PlUl. TrlltlS.
(1799), p. 193. I am indebted to Dr. D. S. L. Cardwell for this point; he has used it in
his teaching to show that even pure 'phenomenological' physics can have its pitfalls.
204 The Achievement ofScientific KnollJledge
theories clutter the teaching of science, not only in the schools, but
wherever it is being conveyed in a standardized form as atool-subject.
But to attempt to clean up all this confusc!d mass, and to purify the
material by deriving it directly from its basis in modem research,
can lead to a syllabus which, however satisfying intellectually to the
teacher, leaves students in a state of confusion worse confounded.
Hence those old objects of scientific discourse which are still useful
for the description of standard facts, will remain alive even after they
are obsolete for the purposes ofscientific inquiry. And because ofthe
complexity ofthe development of those objects, as well as the variety
of teaching situations, it is impossible to ensure that every student is
brought forward in a neat progression from convenient old errors,
towards powerful new truths. More commonly, the different sets
of objects in a family will coexist, peacefully among teachers and re-
searchers who each use only one set, and confusedly among
students. IS
It is sometimes possible for a field to be 'unified', so that its
II In the teaching of elementary chemistry until recendy, chemical combination was
explained by • theory of 'valency' descended directly from the 'dualistic' theory of
Berr.clius in the early nineteenth century, and given • convincing rationale in terms of
electron structure and the electrically neuttal state of. complete molecule. It was then
difticult for students to grasp the Ceo-valence' of organic compounds, to say nothing of
diatomic molecules IS Oxygen gas. Little did we know that the same difficulties were
present in chemistry in the second quarter of the nineteenth century; that they were
partlyresponsible for the decay in theacceptanceof Avogadro's hypothesis; and that our
'basic'theory was the relic of a grand synthesis that had failed. The influence ofa subse-
quent research ttadition on the formulation of. standard fact can be seen in the case of
Boyle's Law. It was originally expressed in terms of a proportionality between the
'spring' of an elastic fluid, and its density; and we may translate 'spring' into 'pressure'
without too much violence to the original. But in modem times 'density' has been
replaced by 'volume't which in the context of the original law is not the fundamental
physical factor involved. It seems likely that when • more general gas theory developed
in the nineteenth century, with volume IS. significant parameter, the version of Boyle's
Law UICd in elementary teaching was modernized, with. consequent loss of clarity. It is
even necessary, on occasion, for the doctrine of • science IS taught to ignore certain
reliable phenomeDa, which may once have been of great interest, in order to maintain
• tidy structure of theory and ezperience. It might be expected that this will occur in
optics, where the student's eye is involved in the experiments. In Oplies, tile Seien&e of
Yilitm, V. Ronchi cites the example of. simple convex leus with an object placed at the
distant focus; the observer should see virtual images behind me lens at an infinite
distance. No one does, but this does not hinder the teaching of simple lens theory. In
more advanced wave optics, one learns of the position ofimages cast by concave spherical
mirrors; experimental tests of these theories lead to the most astonishing results, Done
fitting the theories. Yet the phenomena of highly curved reflecting surfaces were the
IUbject of great interest from antiquity to the seYeDtecnth century, and it.appears that
they bad to be suppressed in the modem period in the interest of. tidy and plausible
aplaDation of refnction and reftcctioa. (See pp. ~ 133-41.)
Facts and tneir Evolution 205
tension with the arrow's point apiDst • plank of wood (making only a slight mark on the
surface), then withdrawn still extended, and released. The arrow will penetrate the
wood; why? Attempts by pupils (and teachers) to solve the problem shows clearly how
their training in elementary mecbanics is confined to manipulation of standard, safe
examples. This thought-aperiment was cited by Leonardo cia Vmci, StleaitmS from tile
Not,lJoob ••• , eel. L A. Richter (Oxford University Press, 1952), p. 66; and Galileo
strunled ftinly to resohe the paradoxical properties of impact forces (see S. Moscovici,
'Ilemarques sur Ia dialogue de Galilie, de Ia force de Ia percussion',RI'DIIe tl'HilloU-t des
Sne.,,, 16 (1963), 97-137.
U E. A. Carlaon, Tile Gme: II Criti&1Il Hillor, (Saunders, Philadelphia, 1966), displays
this process particularly weU. In his Jut chapter, 'Historical Conclusions' (pp. 244-58),
he asserts that the 'beads on a string' version of chromosomes was 'never more than •
straw man'.
23 I am indebted to Dr. D. S. L. Cardwell for this and many other features of the
eYOlution of facts.
M On the dcpneration of the encydopedic tradition in Rome, see W. Stahl, RtnIIIIII
Facts and tneir Evolution 207
frequendy greater than might be considered strictly necessary for
the particular function of the standardized fact; for the natural
pressures of a didactic situation will produce such a tendency. Also,
the person involved in the didactic task of standardizing material
downwards from one level to another will frequendy lack a sufficient
mastery of the material at the higher level for a good presentation at
the lower.
It is easy to single out schoolteachers as the prime targets for a
critical analysis of standardization. Their social isolation from
research activity cuts them off from the~ stream of sophisticated
discussion of the facts and their objects; and in any event their
standardized versions of facts will usually have little similarity to
those used in current research. Hence they are generally forced to
retail standardizations of standardizations, or vulgarizations of
vulgarizations, as the case may be. These inherent limitations of the
schoolteaching situation, along with its function of imparting basic
craft skills rather than 'understanding', must be recognized if there
is to be any fundamental improvement in its quality. However, the
process of standardization of information and facts, like that of tools,
goes on at all levels up to the highest; any publication other than a
pure research report involves standardization. To do the work well
requires a special skill, which is. rewarded in the lasting success
of a classic textbook and its descendants. To do the work badly is
also possible and easy at all levels; vulgarization and incomprehen-
sibility can be found even in the most advanced and specialized
monograph literature.
At some time in the future, it might be possible to enliven the
teaching of science by re-creating parts of the history of which the
standardized materials are the relics. 25 By such a means, students
could share in the excitement ofdiscovery and conflict which attended
all the great achievements, and appreciate that the apparently tidy
organization of the material is only a means of making it compre-
hensible for its function as a tool. But the difficulties in the way of
such a programme must be recognized. It is not sufficient merely to
have good historical studies (of which not many exist as yet); but the
reconstruction of the context of discovery and debate would involve
Sdmee: Origins, Development "nil l"jlumee to tile LIIter MiIltJle Ates (University of
Wisconsin Press, 1962). A modem analogue is discussed by H. Einbinder, Tile Mflh of
tile Brit""";c,, (Grove Press, New York; MacGibbon, London, 1964).
25 A model for such an approach is V. Ronchi, Optics, tile Sdmee ofYisitm, which I
have already discussed.
208 The Achievement ofScientific KnollJledge
teaching some amount of quite obsolete and erroneous technical
material, which could be laborious and also confusing to the weaker
students. It would seem that such an approach to the teaching of
science will be restricted to a very few convenient examples, until
there are deep changes in the tools and techniques of teaching. One
can imagine a separation ofthe two aspects oflearning, with students
mastering manipulative skills privately on a machine, and then
having open-ended discussions with the teacher on the under-
standing of the material at all levels, technical, historical and philo-
sophical. But for this, neither the machines, nor the historical and
philosophical materia1s, nor the t-eachers, are likely to be available
for a long time to come.
7
THE SPECIAL CHARACTER OF
SCIENTIFIC KNOWLEDGE
6 The faith in the essential clarity of mathematics was most beautifully expressed by
Fourier. Speaking of the 'analytical equations' of Descartes, he said: 'There cannot be
a language more universal and more simple, more free from errors and from obscurities,
that is to say more wonhy to express the invariable relations of natural things.' On
mathematics in general, 'Its chief attribute is clearness; it bas no marks to express am-
fused notions.' Tile Aulytical Tlleory ofHeat, 1822, tr. A. Freeman, 1878; reprinted
1955 (Dover Publications) p. 7. A mntrary impression mmes from Karl Menger:
'''Variable'' undoubtedly is among the most frequent nouns in the mathematico-
scientific jargon and hence one of the most successful words ever created•.•• Aetua1Iy,
few scientists seem to have given any thought to the problem of what variables are, and
still fewer of those who use the term formulate clear and satisfactory answers when the
question arises.' See his 'Variables, Constants, F1uents', in C",mt Issues i" tile Pllilosop"y
ofScience, cds. H. Feigl and G. Maxwell (Holt, Rinehart, & Wmston, New York, 1961),
p. 30+ This latter example indicates that mathematics contains obscure objects other
than those which have received attention in the dominant ttadition in 'foundations of
mathematics'.
, The pandoxes of change, and also of aggregation, propounded by Zeno of E1ea,
still exercise philosophers. See H. D. P. Lee,Zmo ofElea (Cambridge University Press,
1936), for tests and aualysis; and for a review of some recent controversies, A.
Grilnbaum, MotlmJ Science and Zeno's P",lIIlozes (Allen & Unwin, London, 1968).
8-s.IC.S.P.
214 The Achievement ofScientific Knowledge
argument, such as 'number', defy all attempts at conclusive ex-
planation. For what we have for analysis are intellectual constructs
derived from very deep aspects of human experience of coping with
the external world. Our practical command of them has developed
through millenia of experience; and no single formalization can
capture that body of completely tacit inherited knowledge.
The ordinary objects of scientific discourse are more specialized
and artificial, and in that measure are more susceptible of intellec-
tual analysis. But as they appear in an argument, they are mixed with
these more fundamental, unanalysable concepts, and it is not
possible to give a clear specification of their meaning relative to
neady defined alternative meanings of the basic concepts of ex-
perience. 8 More important, however, the investigation of the
scientific problem in which they are used or created does not have
the goal of an explication of their meaning. Although the con-
clusion of the argument concerns properties of these intellectually
constructed objects, the properties sought for are those which relate
to experience ofthe external world; for the ultimate goal ofscientific
inquiry is to advance knowledge of that external world. A problem
solved by conclusions about the objects as intellectual constructs
is a philosophical one. The two sorts ofarguments will have different
methods, and admit different sorts of evidence. Although there will
be borderline cases, the distinction between the two sorts of prob-
lem was well expressed, for mathematics, by C. S. Peirce: Mathe-
matics is the science which draws necessary conclusions, while logic
is the science of drawing necessary conclusions. C)
• For example, 'electric current' bas 'operational' definitions through a variety of
effects (the two most important being electromagnetic and elcctrochemica1), each one
of than measured by devices capable of precise specification in terms of elaborated
theories. Although. penetration to the foundations ofany ofthese will reveal obscurities,
these foundations are well hidden beneath the superstructure of accomplished scientific
knowledge required for the framing of the definition of 'current'. It is only in the
determination of the values of the 'fundamental c:oustants' that experimental technique
must reckon with conceptual difficulties.
In his classic monograph, .A" .ACltnml oftile Prifleipks ofM'lIsuremmt lind CllkfJlltitnJ,
N. R. Campbell observes 'the-most acx:omplished physicists are apt to flounder when
plunged into • discussion arising from general principles of measurement; international
amunittees, cbarged with the definition of standards, do not seem (to me at least) to
display that easy mastery oftheir subjects which is to be expected from the conttibution
of their members to original1eaming' (p. vi). For an example of the complications from
which the international committee on electrical units did not quite exttie:ate itself,
see pp. 131-3.
9 See C. S. Peirce, 'The Essence of Mathematics', in ESSIIYs in tile Philosophy of
Snmee, ed. V. Tomas (Liberal Arts Press, New York, 1957), p. 266.
The Special Character ofScimtifi& Knowledge 21 5
school bear the same relation to those ofa single problem as strategy
to tactics in military operations. They require a longer perspective
into the future, and an assessment of more complex and imponder-
able factors; and hence extra qualities of leadership are required for
the effective direction of a school. But in both cases, decisions must
be made for the concentration of resources and work in a particular
direction, to the exclusion of others. In the case of the strategy of a
school, the carrying out of a basic decision will require more time,
and will have greater consequences for the field in which it is
operating. Yet there is the same inescapable element of risk in any
such strategy; the assessments on which the decisions are based are
conditioned by the controlling judgements of value, feasibility, and
cost.
The strategy of a school will have an obvious influence on the
problems which are chosen for investigation; and there is an in-
fluence, less obvious, but perhaps even stronger, on those materials
that are kept alive in the public channel of communication. For,
apart from those results that find a niche in a handbook collection
of very standard information, the material that is chosen for in-
clusion in surveys of a field, and then in specialized and advanced
textbooks, will reflect the judgements of value of those who make
the selections. In this task, part ofthe value ofa piece ofinformation
is its capability for fitting into a coherent account of an important
topic. Stray results which did not derive from problems in the
family worked by some school will be more difficult to fit in neady,
and so will tend to be neglected. Materials which are left out ofsuch
collections are still available, in principle, to those from other fields
who might be interested in using them as information and tools.
But they require the work of retrieval for their discovery, or even
for the establishment of their existence; and the presence of a con-
venient collection, implicidy claiming comprehensiveness for its
functions, will discourage further searching. A similar effect, in the
burying of material that does not find a natural place in a coherent
account, will occur even more strongly in the selections made for
oral teaching. In this way, the advanced training of recruits is
inevitably conservative; by these omissions, their picture of the
field as a whole is restricted to those parts which have been worked
over by leading schools. 19
19 This process of exclusion of the results of apparendy marginal fields can be seen
clearly in the historical literature on lCientific disciplines, which reflects their
226 The Achievement ofScientific Knowledge
The traditions of scholarship within the history of science have
hitherto inhibited the chronicling of the rise and decline of success-
ful scientific schools; but materials are now becoming available for
two illustrative case histories. The first is French physical science
in the early nineteenth century, which provided brilliant results, an
example of a style of work, and a model of excellence for con-
temporaries and successors all over Europe. 20 It is well known that
French influence radiated outwards, first to Germany and then
elsewhere; but general histories have tended to neglect what was
happening Pi France itself after its period of greatness. It now
appears that we should speak not of one cycle but of two; the first
being 'Laplacian physics', and the second, 'anti-Laplacian physics'.
Their creative periods can be roughly dated to the intervals 1800-12
and 1815-23, respectively. The strategy of the Laplacian school was
to apply mathematical arguments to objects of inquiry derived from
an eighteenth-century Newtonian tradition, for the explanation of
old and new experimental results. Its characteristic feature was the
assumption of 'imponderable (and intangible) fluids', consisting of
particles related by short-range centrally-direeted attractive or
repulsive forces, as the agencies of heat, light, electricity and
magnetism; and it was also hoped to explain the laws of chemical
affinity on this basis. The leaders ofthe school, Laplace and Berthollet
had prestige, wealth and influence; and they were able to recruit
the most brilliant students of the Ecole Polytechnique into their
circle, centred on their adjacent residences in the suburban village
of Arcueil. The school enjoyed some striking successes, notably
aeIf-amsciousnes8 and teaching at any given time. Thus, E. Nordcnskiold, Tile History of
Bioloo (Knopf, New York, 1928; original in Swedish, 1920-4), still the classic in its
field, devotes 3 pages out of 616 to ecology (pp. 558-61).
ao The fundamental study on science in Napoleonic France is M. P. Crosland, Tile
Sotiety of ktueil (Heinemann, London, 1967). The distinction between the two
schools, of the Napoleonic and Restoration periods respectively, with the comparison
oftheir styles and fates, has derived from my reading ofa draft ofR. Fox, 'The Rejection
of Laplacian Physics; Turning-Point in the History of the Physical Sciences in France',
to appear in k,law for tM History oftM u",t Seittl&ts. Two other important studies
of the problem are J. W. Herive1, 'Aspects of French Theoretical Physics in the Nine-
teenth Century', BritisA Jowul for tM History of Samet, 3 (1966), 109-32 , and J.
Ben-David, 'The Rise and Decline of France as a Scientific Centte', Mi1JtrV(l, 8 (1970),
160-80. It is only in recent years that historiaDs have perceived that there was a decline;
general histories, either of acience or of French culture, content themselves with de-
saibing the achievements and prestige of French science during its period of greamess.
An enriched perspective on the Arcueil group itself is provided by O. Hannaway in
his review of M. P. Crosland's book, in 1m, 60 (1969), 578-81.
The Special Character ofScientific KnoJl)lttlge 227
6 John Ziman in PuIIlie K1UJ1IJletl", mentions the case of 'continental drift', which
'dropped into the limbo of cranky and speculative notions' for nearly a haIf-eentury
(PP·5&-')·
7 This tendency wiD be most marked in sciences which are not fully matured, and
where" the innovation is related to • folk-science hostile to an academic or 'professional'
science. The persecution of those who used 'mesmerism' (or the relief of pain during
surgical operations in the nineteenth century, is recorded by E. Boring, .A History of
Ezperimmtal Psy,1uJlogy, 2nd ed. (Appleton-Century-Crofts, New York, 1 950 ), Ch. 7,
PP·II6-33·
268 Social Aspects ofScientific Activity
and those whose property is destroyed, are hurt in the process. Two
features ofthis situation make it more serious, in some respects, than
the inherent conflict between general goals and private purposes
which is involved in the normal maintenance of quality control. For
someone who loses property through the rejection of his research
report by the journals is generally at fault for having misinterpreted
the ruling criteria of adequacy and value. It should be part of the
scientist's craft skill to know what sort of result is good enough for
authentication as genuine property. Also, in each such case, only an
individual or a very small group is involved, and their investment in
that particular problem is necessarily limited. Here, on the other
hand, reliable criteria of assessment are not furnished by the craft
knowledge of the field; and the investment that is in danger can
involve the careers of many men.
Yet the continuous advance of science requires that such conflicts
be resolved, and that the community should occasionally decree the
destruction ofthe property ofeminent members, in the service of its
collective goals. How is this to be done? If there were some simple
crucial test for deciding between competing approaches, then in-
tegrity would demand that on some occasions, an elderly scientist
would have to congratulate a young man on just having ruined his
own life's work. As an alternative, one might have an impartial com-
mittee of adjudication, taking evidence from both sides, and then
deciding which strategy is to be supported, and which suppressed.
Such a solution would have the merit of finality, but it is clearly un-
workable. The criteria involved in the assessment of competing
strategies, are so complex, subde and speculative, and the property
at stake is so significant, that for justice to be seen to be done in such
cases, there would have to be developed a system of tribunals, with
their own case-law and intellectually constructed categories of
evaluation. But since there is in principle no simple criterion for
choice in such practical problems, there can be no simple and certain
means to their resolution. In the development of a field as a whole
even more than in the investigation of individual problems, detours
and their associated waste and personal tragedy are inevitable.
Thus, it seems that the dangers that arise from the mismanagement
of novelty are not so much that bad old theories will be retained and
new ones suppressed; for in the long run, provided that there are
some independent centtes even in remote nations, good work will
be recognized. Rather it is that wherever the protection of the
established property of a school becomes too rigid, then in the area
where that school holds political control the stagnation of the field
will be felt by new and potential recruits, and the really gifted young
men who are necessary for the rejuvenation of the field will simply
not be there.
Postscript: Modem Times
Bitter resistance to novelty, like the bitter disputes over priority,
seems to have declined in the recent period. This change is doubdess
due in part to the changing common sense ofscience, with an aware-
ness of the rapid obsolescence of theories and of whole fields. Also,
the change in the location of the scientist's intellectual property,
from the monument of past achievements to the contacts for future
work, enables pastwork to be jettisoned with far less cost. In addition,
there might be a cultural influence, marked by the change from the
German master-scholar and his group of disciples to the more
mobile and egalitarian society of the American model. Because of
these influences, the practical problem ofthe management ofnovelty
is less severe, and the means to its resolution are more effective. The
weaknesses of the present situation are also easy to perceive. Not all
that is new is good, and there are dangers in the dominance of
temporary fashions on scientific inquiry. Problems which require a
long time for their maturing will be less likely to be investigated,
since their results may be considered obsolete by the time they are
achieved. The prestige given to novelty of any sort weights the
scales in favour of the younger men of brilliance, and against the
older men of wisdom. In some fields, it may be quite in order for a
scientist to become an elder statesman at thirty-five; but in others,
where years of experience are necessary before worthwhile new
Laurent: 'I have not been able to dismiss an emotion of indignation in seeing certain
chemists first caD my theory absurd, then much later when they have seen that the (acts
are in agreement with my theory better than they are with all the othen, pretend that 1
have taken some ideas of M. Dumas. Ifit fails, 1 shaD be the author, ifit succeeds another
will have proposed it. M. Dumas hu done much (or the science; his part is sufliciendy
peat that one should not snatch from me the fruit of my labors and present the offering
to him.'
272 Social Aspects ofScientific Activity
results can be achieved, the accent on youth can lead to a loss of the
inherited successful craft experience of the field. These considera-
tions are very subde, but they are relevant to the judgements by
which the leaders of fields, and of science as a whole, distribute the
rewards which control quality and set the direction of research.
10
Up to this point, our discussion has been restricted, for the sake of
simplicity, to 'scientific' problems, where the goal of the work is the
establishment of new properties of the objects of inquiry, and its
ultimate function is the ·achievement of knowledge in its field. For
an historical or philosophical study ofscience in the period preceding
its industrialization, such a restriction is natural; then, it was only
occasionally, and in certain specialized fields, that inquiry of this
sort could be applied directly to the mastery of the natural world.
But with the interpenetration of science and industry, this tradi-
tional separation has come to an end. Indeed, most of the tasks
undertaken by the contemporary corps of 'scientists' have very
different functions, and even different goals as well. The emergence
of this new sort ofwork has led to a variety of new descriptive terms,
which reflect the confusion over its nature and its relations with
science of the traditional sort. The term 'science' itself has stretched
to include technology, and even any sort of inquiry whose methods
are modelled on those of the experimental and mathematical natural
sciences. This multiplicity of meanings is all encased in the single
plausible tide of a projected new field of inquiry: 'the science of
science'. Since these new sorts and extensions of 'science' are those
which present the most severe problems in their relations with the
social and natural environment, we must try to achieve some under-
standing of their distinctive character. I
From one point of view, all the new sorts of 'science' are entided
to that honorific appellation: their work consists of problem-solving
on intellectually constructed objects; and there is no clear demarca-
tion between the methods used in traditional pure science and those
in the most speculative ofsocial technologies. We can, however, make
a useful distinction among the various sorts of problems by means of
I The 'philosophy of technology' is a subject which has attracted relatively little
attention in recent years. The record of a symposium on the subject is published in
Teclmolov "nd Cul'ure, 7 (1g66), 301-40. In this collection of papers, that by Mario
Bunge, 'Technology as Applied Science', bears the closest relation to the problems
discussed here.
318 Science in the Modern World
the classification of final causes involved in a task, that I developed
in connection with the problem of quality control in science. We
recall that the task itself has a goal, which is conditioned more or
less strictly by the function which will be performed by the result
ofthe accomplished task; and this in turn is governed by the ultimate
human pwposes which are expected to be served by the performance
of that function. In the previous discussion, we were concerned with
the need for a harmony between these 'objective' final causes and the
'subjective' ones brought to the task by the scientist: his own pur-
poses, in what he is trying to achieve for himself in the work; and
his motives, which bring him there and set his purposes within the
constraints of the task. For the present problem, we shall con-
centrate mainly on the 'objective' final causes, and by their means
distinguish between problems capable of description as 'scientific',
'technical', and 'practical'.
In each case, we consider the task itself as the investigation of the
problem; the 'goal' of the task is the solution of the problem. We
then ask, to what extent are 'functions' and 'purposes' involved in
the specification of the goals of this sort of task. In the case of the
'scientific' problems that we have been discussing up to now, these
higher final causes are not so influential. The function to be per-
formed by the solved scientific problem, that is the contribution of
new results for the advancement of the field, conditions the work
only in a general way, through the controlling judgements of
adequacy and value. The investigation of the problem may change
course in mid-stream, if a more promising solution appears possible;
and then the solved problem will perform a different function, to
no one's regret. The ultimate pwposes to be served by the
scientific problem are quite remote, diffuse, and unpredictable in
detail; we recall the words of Helmholtz on the radical separation
between the applications of science and the goals of the scientific
endeavour.
By contrast, we describe as 'technical' problems those where the
function to be performed specifies the problem itself. The goal of the
task is fulfilled, and the problem 'solved', if and only if that function
can be adequately performed. A design team that produces a device
that performs a function different from the one assigned, will not
usually be congratulated on its initiative and independence. On
the other hand, in a technical problem the work is conditioned only
in a general way by the purposes to be served. Thus, in a com-
Introduction 319
mercial product, what counts is whether a sufficient number of
people expect the product to serve their purposes, and so buy it;
exactly why they like it does not matter except as a guide to the
advertising.
Finally, there is a class of 'practical' problems in which the goal
of the task, in principle, is the serving or achievement of some
human purpose. The problem is brought into being by the recog-
nition of a problem-situation, that some aspect of human welfare
should be improved. Such problems involve intellectually con-
sttucted objects rather than ordinary 'common-sense', as we see
from programmes and debates on such issues as 'education', 'poverty'
and 'health'. We shall see that large-scale practical problems give
rise to subsidiary technical problems, and perhaps to scientific
problems as well. The fulfilment of its goals in a multitude of in-
dividual cases will usually require a large-scale project, with a
hierarchy of decision and control and a division of labour, analogous
to the projects required for the complete solution of a large-scale
technical problem. 2
By means of this distinction we can consider the similarities and
differences among the three basic sorts of problem, as well as the
various sorts of difficulties to which they are prone, individually and
when they are related in practice. We shall see that technical and
practical problems each have their cycles of investigation, and are
subject to controlling judgements analogous to those of scientific
problems. Even in the case of technical problems, there is no auto-
matic mechanism enforcing quality control on solutions, and one
can speak of a degeneration and even corruption in technology as in
science. Practical problems, and the large-scale practical projects
required for their execution, are by their nature subject to the most
severe hazards and pitfalls. An understanding of these is most
urgendy necessary; for the increasing complexity and sophistication
of our society throws up a rapidly growing set of practical problems.
Even problems previously considered as purely 'technical' are
3 My c1assification of scientific, technical and pnctical problems is closely parallel to
Aristode's classifications of 'theoretical', 'productive', and 'practical' knowledge. See
the Ni'01NU1Ie." Ethi's, vi, 3-5. 1139bI4-114Ias, Met.physits, E.I.1025b, and Sir
David Ross's commentary on the latter text in his edition (Clarendon Press, Oxford,
1924). Aristode distinguishes between 'making' and 'acting', and observes that the
latter is not governed by any end outside itself. This corresponds to my distinction
between 'performing a function' and 'achieving a purpose' for the goals of technical
and practical problems respectively; the function to be performed by a device is always
governed by some ultimate purpose or purposes.
320 S,imte in the Modern World
increasingly being debated in their 'practical' aspects under the
slogans of 'environment', 'pollution', and 'amenity'.
To achieve this understanding, we must appreciate the con-
sequences of the ineffectiveness or immaturity of the disciplines
(usually those concerning man and society) on which the investiga-
tion of practical problems depends. The analysis of the state of
immaturity is not a pleasant task; the condition itself is generally
regarded as at best a misfortune to be remedied quickly, or at worst
as a scandal to be concealed. But without a knowledge of the deep
and systematic differences in the practice of inquiry, and in the sort
of results obtainable, between immature and mature disciplines, we
will have no way of controlling the harmful effects (to themselves
and to society) of ineffective fields grown beyond their appropriate
size and prestige. The analysis of the credence given to the claims of
patendy ineffective disciplines leads us into a discussion of the
category of 'folk-science', which applies equally well to the most
sophisticated academic exercise as to peasant superstitions. With
this battery of concepts we can return to a discussion of the hazards
and pitfalls of practical projects; and see that the wreckage of so
many of the great schemes for human betterment is, sadly, only the
natural state of affairs, at least so long as our conception of 'science'
is that inherited from Galileo and Descartes.
12
TECHNICAL PROBLEMS
20 In his article 'Can Technology Replace Social Engineering ?', Bullet;n of tile
Atomit Stimtists, 22 (December 1966), 4-8, Alvin Weinberg argues cogently that
'social problems can be circumvented or at least reduced to less formidable proportions
by the application of the Technological Fix'; but he is fully aware that 'social en-
gineering' cannot be dispensed with.
21 In 'The New Estate', Bullet;n oftile Ato",;t Scimt;sts, 20 (February 1964), 16-19,
Alvin Weinberg gives a good description of this tendency. 'With respect to the style of
intellectuality, I refer to the tendency to deal with difficult teclmo-social problems-
such as the population problem, or aid to the underdeveloped countties, or the lag in
our civilian technology-by deifying research and denying engineering. I have read
356 Science in the Modem WorltJ
least those of ~ing interpreted by. aggrieved parties as a deliberate
attempt to defer any decision indefinitely.
The function of the conclusions of the phase of enquiry is to
provide a basis for a decision; and this must be capable of execution
as a technical project, by whose means the desired general purposes
may be achieved. The technical project itself must be capable of
being analysed into a multitude of routine tasks, governed by an
hierarchy of decision and control. Further pitfalls are encountered
here; even if the intended functions are appropriate to their ends,
the aggregate of unit tasks, as they are articulated and then con-
trolled by a bureaucracy, may come to be governed by goals which
are contrary to this original function. The operation of overall con-
trol over the phase of execution then constitutes a new practical
problem in itsel~ with its distinct phases of investigation; and the
neglect of this will put the whole practical project in jeopardy. Thus
the inherent complexity of a practical problem, both in its objects
and in the cycle of its solution calls for a diversity of operations,
skills, and approaches even greater than in the case of technical
projects. The application of 'scientific method', not merely as a
crude imitation of the mathematical-experimental sciences but even
in the sense of scholarly inquiry, is insufficient by itself. What is
required is an appreciation of the variety in the nature of the
problems encountered, the difference in their methods, the criteria
of adequacy of their solution, and in their characteristic pitfalls;
and an awareness that there exist some important problems, of
any of these sorts, which are incapable of solution under any
possible circumstances.
mcndation for 'patience and boldness' for this 'infant science', together with a link with
common sense experience. (I am indebted to Miss Helen Gower for this reference.) For
a aitical discussion ofAmerican sociology, in which many of the points made here about
the character and problems of immature disciplines are argued at length and with
examples, see Sociology on Trilll, eds. M. Stein and A. Viddach (Prenti~HaII, 1963).
(I am indebted to Mrs. J. Wootton for this reference.)
It is noteworthy that economics, by contrast, does not have any significant debates over
its methods and their appropriateness. The Festschrift to Oskar Morgenstern, ES$IIYs ill
Mlltllmulticill EctJtUJmics, ed. M. Shubik (princeton University Press, 196'7) contains •
variety of papers on theoretical economics and games theory, but not one discussing the
critical analysis of 0" the ACcurMy of EctnUnlli& ObsertJllhtmS. The reason may be that if
Morgenstern's recommendations were to be accepted, and economists did somehow
learn that a number is an estimate and not a fact, then both research and teaching would
need to be drastically overhauled, and many flourishing lines of research postponed for
at least • decade until worthwhile data could be obtained. But it cannot be said to be a
healthy situation when the published analysis and recommendations of an eminent
member of a discipline are recognized only informally, and ignored in reasearch and in
teaching.
15 In his History of Ezperimmt.1 Psychology, G. Boring identifies four very great men
in psychology's history: Darwin, Helmholtz, James, and Freud (p. 743). Only the last
two had any concern to be considered as 'psychologists', and Freud was, of course,
excluded from academic psychology all his life. Comparing Francis Galton to Wundt,
Boring said ofthe former 'He had the advantage ofcompetence without the limitation of
being an expen' (p. 462).
380 Science in the Modem World
social mechanisms appropriate to a matured discipline, and even
more the task of improving their real condition, are rendered yet
more difficult. 16
Poli-Science
The category of,folk-science' has long been recognized; but it has
always been believed to apply only to preliterate or subliterate
cultures. It is a part of a general world-view, or ideology, which is
given special articulation so that it may provide comfort and reas-
surance in the face of the crucial uncertainties of the world of
experience. In earlier ages, some ofthe leading sciences and educated
arts grew out of popular folk-sciences; among these were the ancient
art of prediction, as well as other forms of magic, along with the very
sophisticated arts of alchemy and astrology. Mer the relegation of
such fields to the status of pseudo-sciences in the seventeenth
century, it appeared to scholars that there was an absolute distinction
in character between rational, progressive 'science', and the super-
stitious and retrograde folk-sciences. It could not be denied that in
the historical order genuine science emerged out of folk-science;
but the causes of this deep ttansformation remained a mystery, and
survivals of the old in the new presented a serious embarrassment.
But the functions performed by folk-sciences are necessary so long
as the human condition exists; and it can be argued that the 'new
philosophy' of the seventeenth century, with its disenchanted and
dehumanized world of nature and its appreciation of closely con-
trolled experience, itself functioned as a folk-science for its audience
at the time. For, as it appeared then, it promised a solution to all
problems, metaphysical and theological as well as natural; and it
Immature and Ineffective Fields ofIntJUiry 387
gained a stable social base long before its achievements were com-
mensurate with any of its claims. 22
Indeed, we may say that the basic folk-science of the educated
sections of the advanced societies is 'Science' itse~ in various senses
derived from the seventeenth-century revolution in philosophy.
This is quite explicit in figures of the Enlightenment such as
Condorcet; 23 and a basic faith in the methods and results of the
successful natural sciences, as the means to the solution of the
deepest practical problems, can explain the absence of genuinely
critical studies of the natural sciences as cultural and social pheno-
mena until very recendy. 24 Experiences which are inexplicable
within the categories of this 'Science' (as those indicating powers of
mind independent of matter) are denied significance or even exis-
tence, for they constitute a challenge to an entire world-view, and
hence threaten chaos and disintegration. A more visible and vulgar
effect of the role of 'Science' as the dominant folk-science of our
time is the status that the tide itself confers on any field of
inquiry or practice. Grabbing for such smtus, with its attendant
social and material benefits, is most widespread in America,
where the folk-science of 'Science' seems strongest and where
control on tides is weakest; 'mortuary science' is a well-known
case of a craft dubbed science, with its attendant academic
apparatus. 25
sa J. Ravetz, 'What was "The Scientific Revolution" ?', ltUlUI" Jowul ofHill., of
Sdmee, 1 (1g66), 15-21.
23 See Condorcet, Espisse tl'un TllbleQu historipe tles /Wo"is tle l'esprit 11"";" (1795),
ed. o. H. Prior (Boivin, Paris, 1938); 'Neuvieme ipoque', pp. 145-202. 'Toutes les
erreurs en politique, en morale, ont pour base des erreurs philosophiques, qui elles-
memes sont li&s l des erreurs physiques. · .' (p. 191).
24 It is significant that through the entire nineteenth century none of the great
sociologists analysed science as a social phenomenon. Comte took his idea of 'positive'
from the successes of the natural sciences; and for Max Weber natural science wu the
purest realization of 'rationality'. The earliest attempt I know at analysing the style and
metaphysics of natural science as the product of • particular culture was that of Max
Scheler, Die Wissmsformm fItUl tlieGesells,IuIji, 1926. See J. R. Staude, Mu S,heler,'"
ItJlelle,tul Portrllit (The Free Press, New York; Collier MacMillan, London, 1967),
pp. 18cH)1. Scheler's insights seem to have derived from his long-standing contempt for
the bourgeois and his belated conversion to a mystical philosophy. There may have been
a current of Marxist analyses already; shortly afterwards quite similar ideas were
expounded by Marxist writers.
25 A list of American 'sciences' has been provided by C. Trusedell: 'A short nndom
search of the catalogues of Graduate Schools delivered, in addition to the ubiquitous
"Social Science", "Political Science," and "Computer Science," some special examples:
"Meat and Animal Science" (Wisconsin), "Administrative Sciences" (Yale), "Speech
Sciences" (Purdue), "Libnry Sciences" (Indiana), "Forest Science" (Harvard), "Dairy
388 Science in the Modem World
The relations between disciplined inquiry and the folk-science
aspects of a subject are naturally complex and variable. 26 To make
the distinction clear, we can consider the case where the same field
could, in different aspects provide both the technical setting for a
scientist's research, and also a folk-science for his personal belief.
This was quite common in the period when science was close to 'the
philosophy of nature' in spirit and name. The quotations from
Helmholtz and Einstein given earlier (pp. 39 and 66) each show
how reflection on some more or less extended part of science, either
in its methods or results or both, could provide comfort and reas-
surance. Indeed, the Helmholtz quotation, and our discussion of
ethics in science, show that without some personal adherence to an
ideal of science as providing some very important knowledge, this
very demanding work could not be sustained at the highest level for
more than a few generations.
The external groups for which a subject functions as folk-science
can vary enormously in their size, sophistication and influence; the
particular folk-science itself can be more or less central to their
ideology, and more or less durable. The style of work in the folk-
science will depend on such circumstances; for a sub-literate
audience it needs fewer trappings of academic jargon and titles,
while a sophisticated audience requires a reasonable facsimile of a
leading branch of'Science', such as physics. But in any case, the clue
to a subject's being a folk-science is that belief in it is quite indepen-
dent of its achievements in producing facts or knowledge, or in
accomplishing genuine solutions of technical or practical problems.
Since the functions performed by a folk-science are so different from
those of the solutions of problems investigated in a disciplined style,
the relevant criteria of adequacy and value of its results will be cor-
respondingly different. Value is determined by the degree to which
a problem-situation is central to the experience of the audience; and
adequacy by the success in offering reassurance and the promise of
understanding. For an educated audience, the mere existence of its
folk-science as a recognized academic discipline may be sufficient to
these ends.
Science" (D1inois), "Mortuary Science" (Minnesota). EsSilYs in the History ofMechanics
(Springer Verlag, New York, 1968), p. 75.
26 A brief but illuminating analysis of the character of the interaction between the
'scholarly' and 'folk' aspects of a discipline, from • Marxist point of view, applied to
Marxism i~ is given by A. Gramsci, 'Marxism and Modern Culture', in his Tire
Motlml Prinee .rul otller "";ti,,,s (Lawrence & Wishart, London, 1957).
Immature and Ineffective Fieltls ofI"'l-Y 389
On occasion, adoption as a folk-science by some influential group
can provide a maturing field with social support (in the way of
prestige, jobs, finance, and recruits) sufficient to enable its firm
establishment; such seems to have been the case with history in the
nineteenth century. Conversely, some particular results ofa matured
field may (with suitable interpretation and vulgarization) suddenly
become a component ofa leading folk-science, and embark on a new
career independendy of the thoughts and desires of their creator;
such was the case with 'Darwinism'. 27 But there can also be conflict,
especially when a challenge to a scientific orthodoxy is made from a
base in a popular folk-science, and uses a style of inquiry and tactics
of struggle appropriate to that base. The ensuing struggles can be
very bitter, for the threat is not merely to aschool, but to the autonomy
of the world of established science in its goals and methods. Even
when a challenge to scientific orthodoxy is made by a man thought
to be a scholarly crank attacking a strong field from a shallow
popular base, the shadow of a threat can be sufficient to cause a
violent reaction; this was seen in the extreme measures taken by the
American community of physical scientists against Emmanuel
Velikovsky and his theories of recent cosmological catastrophes. 28
Immature sciences are even more vulnerable, of course; and (as we
shall see) more closely related to folk-sciences. Hence in them the
sttuggles between competing schools can involve not only doctrines
and personalities, but rival ideologies, their related folk-sciences,
and perhaps their political associations as well.
Folie-Science and Ideological Conflicts
The conflicts between academic science and strongly based folk-
sciences provide many insights into the social situation of science,
27 The interpretation of a chapter in intellectual history in termI of the folk-lCience
functions of an idea has been done by J. w.. Burrow, Ew/lltitnJ tUUl StJtiet,: • SIfIIl, of
VitltwUl" Soeial Tuor;, (Cambridge University Press, 1966). He states, in the Preface,
'It will be argued in this book that the seeds of modem sociological theory were to •
ccmsidenble extent implicit in the doctrines which were becoming current in the 18601,
but they were stifled by the overriding needs of an evolutionism which provided what
the Victorians sought in theories of society' (p. xiii).
On 'Social Darwinism' itse~ the standard work is R. Hofstadter, S«W D"";";"" i8
Ameri,." Thought (Harvard University Press, 1944): for the English scene, see G.
Himmelfarb, D",,,,;,, .utu D",,,,;,,;_ Revollll;tm (OJatto It Windus, London, 1957),
especially cbs. 14 and 19.
28 For a well-argued presentation ofVe1ikovsky's side ofthe story, see Tile YeliitJ'Dsi,
Ajffli" ed. Alfred de Grazia (University Boob Inc., New York; Sedgwick It Jackson
Ltd., London, 1966).
390 Seim&e;" the Modem World
and into some insoluble problems inherent in that situation. Such
conflicts have been little studied by historians from this point of
view, and so are worth some extended discussion here. These
conflicts take two very different forms. The first occurs when the
results of disciplined scientific inquiry contradict the beliefs of a
folk-science, usually a popular one which is also adopted by the
established cultural organs of society. The second sort occurs when
a popular folk-science, perhaps with a radical social and political
message, attacks academic science for its style and connections as
much as for its content. The former sort of conflict is better known;
the history of such conflicts can be read as a series of ttiumphs for
intellectual integrity and freedom ofthought, against popular super-
stition and imposed dogma. Indeed, there was a strong tradition of
such histories, starting in the Enlightenment and continuing through
the nineteenth century; for the sttuggles were real and bitter, and the
camp of'science' stood for the highest values ofcivilized life. 29 How-
ever, the work of science, when it is so engaged, is something more
and something less than that ofa self-contained 'positive' discipline.
There is an added dimension of social and philosophical involve-
ment; and those who inherit the successes of that sttuggle, both in
the knowledge achieved and the secure social position won, cannot
honestly wear the mande of the heroic reformers of the past. On the
other hand, 'positive' science can justly claim autonomy from society
and the state, on the grounds that its particular results are neutral
in their social and ideological effects, while the maintenance
of the activity of research is beneficial in a general and diffuse
way.
If we are to understand the moral problems involved in the
sttuggle for 'freedom in science', we must recognize that those who
attack established or official folk-sciences, are advancing doctrines
that are 'ideologically sensitive': their effects may be disruptive of
the stability of society. They may claim the freedom to make their
attacks, but they must then base their claim on some higher principle.
One such principle is the absolute right ofanyone to proclaim Truth;
another is the ultimate value to society resulting from the free clash
of conflicting doctrines. But to claim the right to utter unsettling
doctrines, while denying any responsiblilty for their effects, comes
perilously close to the arrogation of power without responsibility, a
w".,
at The dusic in this type of historical Hterature is A. D. White, A Hi"., of tile
ofSMt&e .,;" T1Ieoltv i8 C1IrisIerulo&
Immature and Ineffective Fieltls ofIntpUry 391
ttaditional variety of immorality.30 For some generations up to the
present time, most academic natural science has been 'ideologically
neutral'; and indeed this neuttality has been invoked by propagan-
dists for science, as a reason for its complete autonomy of goa1s from
outside interference. The exceptions to this neutrality have been
little studied; one which may be historically significant is the close
association, in the early twentieth ·century, of the science of genetics
with a generally reactionary and elitist folk-science of 'eugenics'.31
30 Galileo early received clear warning of the ideological sensitivity of his Copernican
doctrines from his friend and protector, Maffeo Barberini, later Pope Urban VIII. In
March 1615, Galileo received a letter from Ciampoli reporting a mnversation with
Barberini, advising him not to 'exceed the limits of physics and mathematics', for all
author's assertions on matters in the Scriptures can, on publication, be transformed out
ofrecognition. He cites the example ofGalileo's analogy between the earth and the moon
in the casting of shadows on the surface: 'somebody expands on this, and says that you
place human inhabitants on the moon; the next fellow starts to dispute how these can be
descended from Adam, or how they can have mme off' Noah's ark, and many other
extravagances you never dreamed of.' See S. Drake, Distovtrits.tUl OjMlitnu ofGlIlile.
(Doubleday, New York, 1957), p. 158. There is no evidence that Galileo remgnizcd the
force of this problem up to the time of publication of the DWope tnJ tile TIIHJ Wtwill
Systems. There is, however, an odd passage in·the later DistOflses tnJ TIIHJ N", Stimtts,
which may indicate that after the struggle was over, Galileo finally saw what it was all
about. It mmes when one of the interlocuton praises Salviati (Galileo's mouthpiece) for
a particularly elegant demonstration, and remarks how little such achievements are
appreciated. Salviati replies with some reflections on those who claim to be competent
in a field of study, and yet still publicly deny new truths in spite of beHeving them at
heart, 'merely for the purpose of lowering the esteem in which certain others are held by
the unthinking aowd'. (Edizione Nazionale, p. 204; Dover·ed., p. 169.) This particular
audience is hardly ever mentioned in Galileo's published works; and so one is tempted
to suppose that at that very late stage he had finally discovered its relevance to the
struggle which he had lost. One may put it that he had remgnizcd the category of 'dirty
truth'. Modern students mme to a study of Galileo affair with healthy libertarian
prejudices; for the education ofthose in my tutorials, I put the question, 'Would you try
to prevent someone from giving away mloured mmic:s depicting scenes of sexual
violence against children, at a junior school?' Those who would are then in a position to
appreciate that there was a genuine tragedy in Galileo's conflict with the Church.
By mntrast, Descartes saw the point of ideological sensitivity as soon as he heard the
news of the condemnation of Galileo. He prompdy suppressed his completed work
Le Mtmtlt; and three years later, when explaining this, mentioned 'certain persons to
whom I defer, and whose authority governs my action no less than my reason governs
my thoughts ..• '. It seems clear that he had never previously imagined that there could
be a contradiction between these two governing principles, especially for someone so
shrewd, cautious and honourable as himseU: The aucial text here is the opening section
of the sixth part of the DiseOfls de 111 Mlt1uHJe, in which Descartes finds himse1farguiDg
both sides of a question, without a clear aDSwer coming out of the application of his
Method.
31 Of this particular association between science and social affairs, which was con-
veniendy forgotten during the denUDCiations of Lysenko's campaign against orthodox
genetics, some indications are given by Paul Gary Wenky, 'Nlllw, and PoHtics between
392 Seim&e ;" the Modem World
Even to this day, the social sciences are ideologically sensitive, and
affixing the label 'positive' to anyone of them does not alter this
reality. Also, the problems of 'reckless science' and of the sciences
associated with runaway technology have resurrected old moral
problems, and seem likely to raise ideological issues analogous to
those involved in earlier conflicts with popular or official folk-
sciences.
When academic science finds itself on the defensive against
challenges made by popular folk-sciences (as distinct from counter-
attacks by the servants ofan official orthodoxy), it becomes involved
in the deepest problems of politics; and these show how precarious
is the neutrality and universality of scientific inquiry and scientific
knowledge. For academic science has hitherto inevitably been res-
bicted to an elite leisured class and its occasional recruits; and its
conventions and etiquette are meaningless to the less privileged
sections of society. The institutions of science have usually been so
far removed from the perceptions of the lower orders that both sides
are happily ignorant of its position in a class-divided society. But
there have been occasions when a section of the prevailing academic
science has been attacked, for its content, its style, and its social
position, by representatives of a radical movement demanding its
replacement by a populist folk-science. The Paracelsian tradition in
early modem Europe provided a basis for such challenges, of which
the best known is that of the 'sectarians' during the English Civil
War. 31. A similar pattern can be discerned in the attack on the
Acadbnie des Scim&es during the French Revolution;33 and again in
the career of the Soviet agronomist Lysenko. In each case the chal-
lenging folk-science related to a 'romantic' philosophy of nature,
combining a stress on crafunan's manipulation, a personal involve-
ment in the work, a democracy of participation, and a disttust of
the Wan', N""t, U4 (1g6g), 4M-72. I am indebted to him (or this reference; he bas
work in preparation which will show the connectioD fully.
Ja For the English Civil War, the semiDaI paper is P. M. RattaDsi, 'Paracelsus and the
Puritan Revolution', AmlJiz, I I (1963), 24-]2. The theme is developed further by the
lime author in 'Politics and Natunl Philosophy in Civil-War England', Anes ~u Xl-
C"""is l",trUritnuU tl'HiSl,;,e des Stinltes, ii (0sI0Iineum, Warsaw, 196'7), 162-6. The
basic printed clocumentl (or this CODttoYeny are the polemical worb on the universities,
publisbed in the early 16501; see A. G. Debus, Stinlte IIfIIl EtlwMitnl i8 ,lie Stw1JIemlll
C""",: ,lie Wei",,-W.4 DeNte (Oldboume, London, 1970).
JJ For the ideological coo8icII over ICieDce in the ndicaI phase ofthe French Revolu-
tion, lee C. C. Gillispie, 'The E'"1e1tJliilU and the Jacobin Philosophy o( Science: I
Study in Ideas and Consequences', in CritieIJI Pr,i1nlu i8 ,lie Hi".:! ,f Stinlte, eel.
Immature and Ineffective Fields ofInquiry 393
abstract or mathematical reasoning. Viewed from the point of
34
Conclusion
From our analysis, it appears that the condition of ineffectiveness
or immaturity in a field is not a simple absence of some desired
characteristic of a matured science. Indeed, a field can be kept in
such a state by inappropriate methods, adopted partly to achieve a
state imagined to be that of a mathematical-experimental natural
science, and partly to pretend that it is already there. Under such
circumstances the very real difficulties of achieving worthwhile
results, of teaching, and of operating the social mechanisms for
the guidance and control of research, are made still more severe.
However, the external constraints on such a field, in the present
social arrangements of science, would make it very difficult for the
leaders ofan immature field to adopt methods appropriate to its con-
dition, recalling the old distinction between 'philosophy', 'history',
and 'art'. When immature and ineffective sciences are enlisted in the
effort to resolve practical problems, their difficulties are still further
compounded. In addition to the hazards of hypertrophy and conup-
tion, an immature academic field is particularly liable to the com-
plex influences from its associated folk-science; and to become
submerged in cross-currents of political and ideological conflicts.
The picture is a gloomy one, but since immature and ineffective
fields are due to be involved in public affairs to an increasing extent
as our social problems become ever more complex, an awareness of
their limitations is necessary if their application is to systematically
produce more good than harm.
Also, with an appreciation of the naturalness of the state of
immaturity, we can see it as one phase ofa full cycle of development,
rather than imagining that every science emerges from a futile pre-
history, onto a permanent plateau of the 'positive' state. With the
state of immaturity corresponding to the infancy of a field, from
which only the hardy survive, we can see the phase of maturity
giving way, in its tum, to one of senescence. The lasting achieve-
ments of the phase of immaturity will be a few aphoristic insights of
a philosophical character, and perhaps some tools and techniques;
those of the phase of maturity will be a mass of positive knowledge.
But when the basic insights are exhausted and the leading problems
shift elsewhere, the field enters senescence, useful only for its
402 SGimte;" the Modem WorllJ
standardized information and tools purveyed through teachers. 43
This cycle is easily observed on the small scale, in descendant-lattices
ofproblems and their associated schools; but it may also apply, on the
large scale to whole sections of the world of scientific and scholarly
enquiry. To be involved in a field just entering maturity is the most
rewarding career for a scientist; for then one can make great achieve-
ments at relatively little risk. But estimating the points of transition
between phases is a very delicate task; a field or area ofscience which
is approaching senescence is a dreary place; and immature fields
with the hope of imminent maturation are, with all their attendant
ha2ards, the place where the greatest challenge is found.
4J The senescence of academic physics has been indicated by the distinguished
physic:iat, A. B. Pippard: 'But we should not forget ~t the great en of academic
physics may well be drawing to its close, and that we are very likely entering a new en to
be domin'ted by fundamental advances in molecular biology, biochemistry and pharma-
coIou, all of CDOI'IDOUI industrial potentiality.' The last phrase indicates that the
eenesceDCe may extend to all of academic science, although the context of the remarks is
the problem of scientists in industry. See A. B. Pippard, 'Innovation in Physics-based
1Ddustry', Alma D of'Tbe Swum Report', pp. 10C>-3. I am pUeful to Dr. J. Wootton
f. this refereDc:e.
Part V
CONCLUSION: THE FUTURE OF
SCIENCE
CONCLUSION: THE FUTURE OF
SCIENCE
Recapitulation
Our analysis necessarily started with an absttaetion from a com-
plex reality: considering the investigation of a problem as the unit-
task of scientific inquiry. We saw that this work comprises several
distinct phases, each involving sophisticated craft skills. Contact
with the external world is made in the production of data; but this
must be converted into information, and then used as evidence in
an argument. The argument concerns artificial objects, intellectually
consttueted classes of things and events; and it is about these that
14-&U.P.
406 Ctmt/usion: The Futur, ofS,ien,e
the conclusion is drawn. Since no argument in science can be
formally valid, and hence no conclusion necessarily ttue, the
acceptance of conclusions must be governed by criteria of adequacy.
These impose a complex and subde sttueture on the argument, and
ensure the avoidance ofthe known pitfalls which can be encountered
in manipulation or in inference. They belong to the body of craft
knowledge of the methods of the field, along with particular tech-
niques of using tools and with other conttolling judgements such as
those of value. The methods are informal and even tacit, and are
transmitted interpersonally rather than publicly; they are incapable
ofbeing tested scientifically themselves, but arise out ofthe collective
craft experience of the field.
The conclusion of an adequately solved problem is still far from
being knowledge or even a fact. The research report on a solved
problem must be assessed by a referee before it is certified through
publication in a recognized journal. Even then, it must prove its
significance (by being put to use), its stability under testing and
repetition, and its invariance under the changes in conceptual
objects which inevitably occur as new problems are investigated. Of
all the facts which are so established, the great majority remain
within the descendant-lattice of problems deriving from their
original, and sink into oblivion when that field is exhausted and
forgotten. Those facts which survive to become scientific knowledge
have a different path of evolution. Rather like successful tools, they
are also extended to other fields, in standardized versions performing
a variety of functions and taking diverse forms. When such facts
have survived the demise of their original problem and its descend-
ants, and remain alive through their many uses, they are recogniz-
able as knowledge. It is paradoxical that the different extant versions
will be incapable of being reduced to a single, standard statement;
and that the obscurities latent in the original formulation will
frequendy remain unresolved. It is also paradoxical that the whole
process of evolution and selection is accomplished by fallible in-
dividuals, and governed by a craft knowledge of methods. Such a
conception of the nature and origins of genuine knowledge runs
counter to the hitherto dominant ttaditions of the philosophy of
science and epistemology. But in them, the basic problem was how
an individual could quickly achieve ttuth or the best substitute.
Here, the guiding principle is 'veritas temporis filla'; and as the
daughter of time, transformed and tested by a great variety of
Co",lus;on: The Future of S,ien" 407
contacts with the external world, recognizable scientific knowledge
emerges from a complex and lengthy social endeavour.
For simplicity this first analysis was restticted to matured fields
of 'pure' scientific inquiry, and it presupposed the presence of
social mechanisms whereby the private purposes of individuals
would be harmonized with the collective goals of the endeavour. An
examination of these mechanisms was necessary as a preliminary to
any analysis of the conditions under which the health and vitality
of science can be maintained. We saw that in the protection of the
intellectual property embodied in an authenticated research report,
the inherited formal system of journals and citations must be
operated by an informal etiquette if it is not to be abUStd and
destroyed; and the inttoduction of new forms of property lacking
the controls of the journals requires a most refined etiquette if they
are not to lead to a degeneration of the work. The management of
novelty has in the past presented some of the most severe practical
problems for science. We saw that neither the assimilation of old
materials to new, nor the choice between competing research
strategies, can be accomplished on the basis of general rules; but
that destruction and conflict are inevitable in science, and the social
organization and style of work in each particular area will determine
whether the outcome of a 'revolution' will be renewed vitality or
stagnation. The social task of quality conttol in science bears most
directly on ethics, for the immediate private purposes of most
individual agents are served by skimping, however slightly, on the
quality of their accomplished tasks. No formal system of imposed
penalties and rewards will guarantee the maintenance of quality, for
the tasks ofscientific inquiry are generally too subtle to be so crudely
assessed; nor will the advantages to an individual of a good reputa-
tion ofhis group be sufficient to induce a self-interested individual to
make sacrifices to maintain it. Only the identification with his col-
leagues, and the pride in his work, both requiring good morale, will
ensure good work. Science possesses a hierarchy of quality control,
informal except at the lowest level where research reports are
assessed for publication; and the conttollers are conttolled by
rewards of prestige in various ways. At the top of the pyramid of
conttol are the leading scientists at the leading universities, who
conttol each other and their fields by the most informal oftechniques.
They are neither answerable to their inferiors, nor sttictly account-
able to those who provide their support; and their work of direction
408 Conclusion: The Futur, ofScience
and control requires both wisdom and the highest ethical com-
mitment if it is not to degenerate. The ideology and social context of
this ethical commitment are inherited from an earlier age; and the
conditions of industrialized science present them with problems
and temptations for which their inherited 'scientific ethic' is totally
inadequate.
Science becomes directly involved with society at large when it is
applied to the solution of technical problems, involving the pro-
duction of the means for the performance of a function, or practical
problems, involving the achievement of the purposes of individuals
or groups of people. Each of these other sorts of problems have their
characteristic cycles of investigation, their appropriate criteria of
quality, and their particular pitfalls. The most sophisticated technical
problems, in which the scientific component is sttongest, are en-
countered in those industries where the automatic mechanisms of
quality conttol through a competitive market are weakest; in them
there is the danger, not merely of runaway technology advanced
without regard for human welfare, but the corruption of the activity
of technical problem-solving itself. The investigation of practical
problems, and their solution through large-scale practical projects,
encounters every pitfall of scientific and technical problems, and
then some peculiar to itself. Conflicting ideologies and purposes are
at the heart ofevery urgent practical problem; they lack the accepted
criteria of quality for their solution; the sciences involved in them
are usually immature; and in their execution they are prone to
distortion by the natural tendencies of bureaucratic operation.
Because of the increasing recognition of new practical problems,
immature sciences are assigned tasks which they are not strong
enough to accomplish properly; to their internal difficulties (ag-
gravated by the necessary pretence of maturity) are then added those
ofhypertrophy. When involved in the solution ofpractical problems,
they also function as folk-sciences; and the resulting confusion of
the different sorts of problems and their appropriate styles of work
can result in total demoralization and corruption. These difficulties
and dangers are directly relevant to the future health of the ttadi-
tional, established, matured natural sciences. For their internal
difficulties of recruitment and morale, as well as those of their
relations with society at large, are practical problems of the sort
analysed here. Their resolution is urgent, because of the delicacy
and wlnerability of scientific inquiry; but the pitfalls to be en-
Conclusion: The Future of Scien" 409
countered here are no less dangerous than on any other practical
problem.
Science in History
Our analysis of genuine scientific knowledge showed that it is the
outcome of a lengthy process operating through history; and we
have also indicated various ways in which scientific inquiry as a
whole is an historical phenomenon: c;onditioned by its social and
cultural environment, and subject to cycles of growth and decline.
The very long period of the flourishing of the matured natural
sciences was plausible evidence for the comforting beliefthat science,
in its academic and positive period, had truly reached evolution's
end, and would thenceforth experience a simple progress onwards
and upwards indefinitely. The 'idea of progress' with which the
rise of modern science was intimately associated received its mortal
blow in the First World War and its aftermath; but in science
itself the assumption survived for nearly another half-century.
Appreciating that that long 'golden age' of science is now definitely
ended, we can see it as one temporary phase in the history of man's
attempts at understanding and control of the perceptible world
around himself. Extending back to remote antiquity, this history has
the common pattern of continuity and change, and gains and losses,
as its successive phases appear. Seeing ourselves in a new phase of
this history, we can face its problems as inherent in its conditions,
and not as merely accidental difficulties to be removed by exhorta-
tions or by administrative devices.
While recognizing the novelty of the problems of industrialized
science, we would be quite mistaken to think that the whole social
activity of science has been completely ttansformed in the last two
decades. A large part of scientific research proceeds as before, in a
social context which is still mainly 'academic' rather than industria-
lized. The radical difference is that certain new tendencies resulting
from industrialization are developing rapidly, and that the self-
consciousness of science, as reflected in the pronouncements of its
leaders, has changed from a simple optimism to a ttoubled uncer-
tainty. Even if these developments continue and intensify, one
cannot predict with any assurance just how serious their effects will
be within any given time. The work of scientific inquiry is now
embedded in a very large social institution which performs other
410 Conclusion: The Future ofScience
essential functions in an advanced society, including higher
education and the investigation of technical and practical problems.
Even ifall these sectors encounter increasingly serious problems they
will continue to receive support from society so long as they are
considered as performing these functions better than any feasible
alternatives. Historical change can take a long time to work itself
out; and we must avoid the naive rationalism characteristic of
radical reformers, who believe that once the insoluble problems of an
institution have been exposed it will soon pass away. We might recall
that the Catholic Church was conscious of a deep internal crisis as
early as the later twelfth century; the incompetence and corruption
of the clergy had already led to powerful movements of reform
and schism. Yet several centuries were to pass before 'Reformation'
achieved a permanent, independent base; and the Church
survived that, to continue as a powerful force up to the present.
Of course, the Church had coherence, wealth, and power in a
way that the social institution of science does not; but if we
are assessing the prospects for science over a period of a few
decades to come, we must keep in mind the enormous inertia of
any established institution in all but the most revolutionary of
contexts.
We must also remember that the world of science is a very varie-
gated one. Some fields are capital-intensive, and so very vulnerable to
the effects ofindustrialization; while others can produce outstanding
work with small investments for each project. Again, some are
closely related to technical problems, while others can proceed in
peaceful uselessness. National styles in science, which were very
marked even among the successful nations before the complete
domination of academic science, may emerge again so sttongly as
to condition the sort of work which is successfully done in different
places. It is well known that the greatest sttength of America lies in
technical problems: both the development of physical devices and
the organization of work and management. But American science is
particularly prone to the dangers of enttepreneurial, shoddy, and
dirty science. By conttast, British science is sufficiendy small and
poor (in comparison) to be led by an institution (the Royal Society)
still resembling a club; and in this context the problems of con-
ttaction, and accommodation to industry, may be managed with
more finesse. Again, in the Soviet Union the political pressures on
intellectuals are so crude, that the leaders of science have a natural
Ctmtlusion: The Future ofScience 411
social and political role as spokesmen for an Enlightenment in
classic eighteenth-century terms. I
The history of science provides yet another caution against over-
simple predictions of the effects on science of the changes in its
context. While it is relatively easy to give plausible explanations of
the gross features of scientific activity in terms of its social and
cultural environment, this becomes progressively more difficult as
one tries to account for work of lasting quality, and the productions
of genius. We have mentioned earlier that some of the greatest
scientific work of all time was conducted within a metaphysical
framework which would now be rejected as superstitious and anti-
scientific. Yet even those who were searching for the divine har-
monies of the celestial motions, or for the material location of the
world-soul (as Kepler and Gilbert, for example) could, by talent and
discipline, achieve results which became incorporated into the body
of genuine scientific knowledge. Similarly, although men of ability
will generally do better work when they are part of a vigorous com-
munity, enjoying prestige and leisure for their researches, some ofthe
greatest advances have come from men working under difficult con-
ditions in nearly complete isolation: such men as Copernicus,
Mendel, Galois, and Lobachewski are cases in point. Hence, even
if the goals of 'positive science' are totally displaced in scientific
inquiry, and the major communities of science experience crises of
finance and morale, there may yet be scientific achievements which
will last for centuries to come. However, there is no known means of
encouraging genius through adversity; and it would be dangerous in
the extreme to conclude that the quality of scientific work would
improve if scientists were left to starve in garrets.
Science in Society: the Problems ofMorality
With all due caution in the face of the complexities of historical
change, we can proceed to indicate the deepest problems that
science will face as the process of industrialization develops, and to
speculate on how they might be resolved. For this we will first need
to analyse the inherent tensions in the relations between scientific
inquiry and the society at large which supports it.
• See A. Vucinich, 'Science and MonJity, a Soviet Dilemma', Snmee, 159 (15 March
c:urren.t movements for autonomy of science, in
1968), 1208-12, for a discussion of the
the context of the Russian traditions. For an example of this style of work, see A. D.
Sakharov, Progress, CtNzistmee "rul ["tellen"," Free__ (Norton, New York; Deutsch,
London, 1968; Penguin, London, 1969).
412 Conclusion: The Future ofScience
Ofitself, scientific inquiry is not a self-sustaining social institution;
neither wealth nor po,ver are derived directly from the activity. It
requires support from society at large, or at least some wealthy and
powerful section of it, if it is to exist. From lay supporters, then,
science requires first of all 'legitimization', if its practitioners are not
to be relegated to a despised or abhorred fringe ofsociety. More than
mere tolerance is required; resources must be invested in scientific
work, both in providing paid research time for the practitioners, and
in supplying their specialized equipment. Finally, an increasing flow
ofrecruits, drawn from the adolescent population, is necessary if the
work is not to stagnate and die. In exchange, science can offer the
promise of assistance in the solution of technical and practical pro-
blems. This may be direct, or indirect; thus high-level teaching,
which in recent generations has been considered to depend on an
association with research, contributes to technique; and the con-
tribution of science to national prestige, or to the strengthening of
the nation's official ideology, helps in the solution of practical
problems. But we notice that these return offerings of science to
society are not, and cannot be, dominant components of the general
goals of the work, and still less of particular research projects. Some
of them may well be present, in varying degrees, in the work of
particular individuals or schools, especially in the conditions of
immaturity and in the endeavours of a genius. But should a large,
established field, depending on the efforts of many research workers,
allow its criteria of value (and hence of adequacy as well) to be
dominated by such external functions, the work which results will
not be science. It may have excellence of a different sort, or it may
be quite corrupt, depending on conditions; but it will contribute to
the advancement of knowledge only very incidentally.
Thus the social position of science is really quite precarious.
Scientists are not professionals ofthe ttaditional sort, who can justify
their position by the serving of the purposes of clients; nor can
science be conducted on a large scale in a social context analogous to
that of the fine arts, providing prestige to particular pattons. Science
not merely requires very tangible support in return for quite intang-
ible returns; but the different components of its support will in
general derive from different sections of society. Each of these will
need to be furnished with propaganda appropriate to its tastes; and
this internal complexity, together with the great variety of social con-
texts within which science has operated, have produced a great
Conclusion: The Future ofScience 413
variety of themes in the literature of justification and defence of
science. The dominant themes from earlier times relate to the
ideological functions (and dysfunctions) of science or of natural
philosophy; for until the middle of the nineteenth century only a
very few fields (as chemistry, itself only recently established as a
science rather than a craft) could make any plausible claim to con-
tribute to industrial production. In the later nineteenth century an
accommodatio' i with industry was recognized as necessary; and in
the present period there is a strange mixture of 'images of science'
purveyed to its different audiences.
In earlier times, the principal threats to the autonomy of science,
or rather natural philosophy, have occurred when some fields were
considered as ideologically sensitive, endangering the established
religion; and in this respect they were involved in the practical
problems of their age. Those with responsibility for the spiritual
welfare of the lay public would use all the means at their disposal to
contain or eliminate the dangerous doctrines and their perpetrators.
Such measures ,vould be more successful in places where the Church
had an established machinery for handling doctrinal crimes, and the
power to enforce its decisions; hence the Catholic Church has had
an unfair reputation of outstanding enmity to free inquiry. As a
result, there developed a belief in a close association between
scientific inquiry and independent, rational, or free thinking in
general. Propagandists for any of these ttaditions have assimilated
the martyrs of science (most notably, the very complex figure of
Galileo) to their cause; and some important traditions within the
folk-history of science have imagined the community of science as
necessarily composed of individuals who are selflessly and fearlessly
devoted to Truth, against Authority and Superstition.
This identification rests on several basic fallacies. Scientific inquiry
must have a subtle and complex assessment of the strength of
evidence deriving from accepted authority; and in this it is similar
to any other work where partly new problems are being solved. Only
in sectarian religion and in teaching can the work continue success-
fully for any length of time without encountering the problems of
the management of authority; and of course the total rejection of
authority, whereby every assertion must be examined as an equal
claim to truth, quickly yields chaos. Also, scientific inquiry is
ideologically sensitive only accidentally and occasionally. In England,
for example, the propaganda for science, purveyed from the
4 14 Conclusion: The Future ofS,;mee
seventeenth through the nineteenth centuries, argued that the
contemplation ofGod's creation could not but induce to ttue religion.
Of course, in England hardly anything gets really sensitive ideologi-
cally; and English divines could use the persecution of Galileo as
evidence for the wickedness of Rome. And when English scientists
were confronted with the practical problems of uncomfortable
theological implications of their work, they were far more likely to
devote their energies to a reconciliation than to make some specialized
results a fulaum on which they would move all heaven and
earth.
Towards the end of the nineteenth century, the applications of
science to technical problems were increasing in number and in
power. Spokesmen for science had the delicate task of securing ever-
increasing support from the community on the basis of such useful-
ness, while still preserving the autonomy of science itself. We recall
the brilliant speech of Helmholtz (quoted on p. 39 above), where he
reminded his audience of the apparent 'uselessness' of Galvani's
experiments on animal electricity, which yielded the eleCb"ic tele-
graph; but then warned them that the scientist himself must not be
expected to search for anything but new knowledge about nature.
This sincere plea for the social support of science, on the grounds
of its accidental social benefits, is characteristic of the period of
matured academic science.
With the advent of industrialization in science the claim of the
technical applicability ofscience in general did not need to be pressed
(although particular fields still need to justify their requests for
support in these terms). Although there have been continuing dis-
cussions about the proportion of resources which should be devoted
to 'pure' or 'basic' or 'undirected' research, there is a general
recognition among policy-makers of science that it performs a
variety of useful functions and so deserves some minor share of the
budget. However, another audience has suddenly become crucial for
the continued well-being and expansion of science: its potential
recruits. Among them, there is a significant fraction who see the
applications of science quite differendy from the nineteenth-century
optimists such as Helmholtz. Not the telegraph, but the Bomb, has
become the type-example ofthe leading technical problems in which
science is engaged. Hence a new, negative, and defensive theme has
been developed in the propaganda for science: its neutrality. Of
course, this will be purveyed to lay and juvenile audiences, in the
Conclus;on: The Future ofScience 415
hope that they are unaware of the firmly realistic terms in which
'science policy' is now cast.
Since the claim of 'neutrality' is the last defence against recogni-
tion of the political and moral problems of science consequent upon
its industriali7~tion, we can expect it to be advanced for some time
to come; and it is worth closer analysis. The attempt to disclaim
moral responsibility for the effects of scientific work has been made,
not merely for 'pure' science, but even for work on technical
problems with a military function. This extension of the domain of
moral isolation of science is too implausible to survive. To be sure,
one can agree that in one sense 'it is the height of folly to blame the
weapon for the crime', Z if for nothing else than that inanimate things
are not appropriate objects of moral judgements. But those who are
engaged in making weapons in the knowledge that their main
function will be in the commission of a crime are subject to moral
judgements and sometimes to legal proceedings as well. If this were
not the case, then the defence of Adolf Eichmann, that he was
merely engaged on a technical problem oftransport, indifferent to the
fact that it was a one-way transport of Jews and others to the gas-
chambers, would be a valid one.
It is more plausible to assert the neuttality of science for that work
which is governed by purely internal criteria of value, so that in the
choice of problems the possible technical functions are either un-
known or irrelevant. Even here, there is an area of ambiguity; for a
particular research worker may choose to work on a problem for its
functions in the advancement of the field, and for its political
functions for his own career, while being aware that the investing
agency is interested in the problem for its possible technical func-
tions. If these technical functions are morally dubious or bad, can
he claim immunity? The ignorance of consequences is not always a
valid defence in law, and the mere absence of deliberate intent to
malefaction is an even weaker defence.
However, there are severe difficulties in the way of making precise
and fair moral judgements on scientists, individually or collectively.
If nothing else, our experience of these problems is extremely short,
and we possess neither general principles nor case-law for their
interpretation, whereby responsibility and blame can be assigned in
any but the most blatant cases of dirty science. Also, the division of
2 See Sir Peter Medawar, 'On the Effecting of An Things Possible', Presidential
Address to the British Association, 1969, NnIJ Seimtisl (4 September 1969), 465-7.
416 Conclusion: The Future ofScience
labour in large-scale technical problems is extremely fine; so that the
scientist who publishes a result generally has no more knowledge of
its possible functions than does a process-worker assembling a
standard component of a device. This does not mean that the posi-
tion of the agent is one of moral neutrality.; rather that it is morally
indeterminate. However, to the extent that his research is related
to technical applications, the area of indeterminacy decreases, and
the scientist's responsibility becomes defined. Once the scientist is
aware of the likely consequences of his work, his sole disclaimer of
responsibility can be along the lines of, 'I was only following orders.'
This is no longer likely to be acceptable as a defence, in science as
anywhere else.
These considerations apply strongly to scientists who are em-
ployees of a 'mission-oriented' research establishment; but for those
whose experience is in a community of science still enjoying an
academic style these moral problems are remote and philosophical
rather than immediate and personal. There, research problems
and personal achievements are evaluated by internal criteria, re-
gardless of the motives of those supporting the work; the hard work
and strict self-discipline are supported by a refined ethic; and results
are shared with colleagues independendy ofall the boundaries which
divide mankind. There, it seems, worthwhile work can be done,
insulated from the moral squalor of ordinary life.
But even to the extent that the moral neutrality of academic
science is real, it creates moral problems on its own; and the deep
connection of science with the culture in which it is embedded
involves science in its basic problems of justification and survival. The
practical irrelevance of most of the results of scientific inquiry is a
blessing to some, but an agony to others. What can be more selfish
than to tum one's back on the sufferings of humanity, devoting
one's talents to narrow tasks which will bring immediate rewards to
oneself and only the most remote and unknowable benefits to one's
fellow man? For a young scientist with a strong social conscience it
requires an extremely strong faith in the human value of scientific
knowledge to justify such a career. 3 Once this moral dilemma is
recognized, it can be alleviated by various sorts of good works, but
3 J. G. Crowther, Tile SodtllReliltiom ofStimte (MacMillan, London, 1941), has some
sharp words on this problem: 'Young scientists who abandon science for politics often
prove to be mentally unstable, and after a few years of bohemian agitation become
conspicuously conservative. Conduct and opinions that appear to be based purely on
moral sentiments are nearly all mspect' (p. 644).
Conclusion: The Future ofScience 417
never completely resolved. Moreover, the remoteness of academic
science from human concerns is itself a result of its own traditions,
deriving from its particular niche in a particular society. The openness
and internationalism of science are admirable and valuable in them-
selves; but they are not the same thing as a sharing with all humanity.
Rather, they are methods of social behaviour of a small group
operating within European literate culture. Now that national
boundaries within that culture are of decreasing emotional signifi-
cance, the ttanscending of them is correspondingly less impressive.
And to the extent that this culture as a whole is subjected to moral
judgement, for its involvement in 'various sorts of colonial and class
oppression, and for its creation of a runaway technology, academic
science will be inescapably implicated as well.
It is quite likely that those of the present generation of elder
statesmen ofscience who invoke its 'neutrality' are trying to reassure
themselves as much as any audience of potential recruits. Con-
sciousness always lags behind reality, and its adjustments are usually
abrupt and painful. The conception of science as an essentially
academic enterprise, and of 'the scientist' as an academic researcher,
has persisted unchallenged in all the literature about science until
very recently, in spite of the ttaditional industrial connections of
several major fields (particularly chenlistry, but also physics and
biology), and in spite of the fact that the great majority of those who
have ever earned their living through their scientific skills, have done
so in technical work. 4 Even the interpenetration of science and
industry can be traced back to the later nineteenth century, and can
be seen as growing continuously since then. An awareness of a new
condition of science came only when a series of dramatic events,
such as nuclear weapons, and new technical problems, such as the
planning oflarge-scale scientific research, obttuded themselves. The
self-consciousness ofscience is still trying to cope with these changes;
the purely technical problems of decision and control are difficult
enough in themselves, and the deeper practical problems of re-
sponsibility and morality are only beginning to be grasped.
In the short run, the easiest response to such problems is to hope
4 N. D. Ellis, in The Snmtifi& Worltr, has shown that the Royal Institute of Chemistry
and the later Institutes of Physics and of Biology were created with the co-opention of
the major employers of scientists; and their ethical principles were framed for the
'professional employee'. The academic scientist was not their concern; and these
institutes are now subject to some strain because the majority of 'Q:S.E.'s' in industry
have no professional status, but are nther 'scientific workers'.
4 18 Conclusion: The Future ofScience
that they will go away, and in the meantime to try to get the best ofall
worlds. Such a course of action, which was almost certainly not the
result of a conscious policy, was taken by British university science
teachers in the post-war period. Still believing in science as a
genuinely liberal education, they expanded their departments with
State funds intended to provide more units of trained manpower, and
in practice taught Honours degree courses designed for that small
fraction who could proceed to research. The instability of such an
arrangement was revealed after only a decade or so; and whatever the
outcome of the pressures for its alteration, the world of science has
suffered no pub.lic discredit thereby. But the same sort ofconvenient
wishful thinking, applied to the understanding of science as a whole,
can have dangerous consequences, including a corruption of the
whole work.
The sort of corruption which can occur in science has little in
common, superficially, with that which is recognized in public life.
Hence it is necessary for us to analyse the concept briefly, to show
why it is relevant to the problem of science. s We can say that an
activity is corrupt when the actual goals ofthe tasks accomplished are
contrary to the professed social functions to a degree that a public
trust is beuayed. Corruption is occasionally flagrant, but more
commonly it exists in a penumbra ofambiguity; both the divergence
between the final causes, and the awareness of that divergence, are
ill~efined. Because of this, a man may work in a corrupt situation
without himself being corrupted. 6 If he is ignorant of the state of
5 The only worthwhile analytical study of corruption is The Autobiography ofutU:ol"
SteJfnu (Harcourt, Brace, New York, 1931). Steffens was able to gain experience of
corruption in American public life through his work as a journalist; but he could rise
above mere 'muckraking' because he had previously spent years of (informal) study and
thought on the problems of ethics. Although he never achieved a coherent solution to the
problem of corruption (which is an interesting problem because, as Steffens found, the
c:orrupton frequendy have more honesty and personal integrity than muckrakers and
reformers), his book is a mine ofexperiences and insights. My attempted formal defini-
tion derives from his interview with 'a dying boss', where Steffens told him why corrup-
tion is evil (see Part III, elL VIII, p. 419).
The problems of corruption in post-colonial societies are discussed by S. Andreski,
Tile Afriea PrIllU.",tJII (Micbael Joseph, London, 1968). Although the author discusses
the problem with sympathy and insight, and even ofFers the technical term 'kleptocracy',
his analysis Iacb the depth ofSteft"eus' (ofwhich he seems unaware), and it suffers from
his assumption that in advanced societies corruption is minor in its scale and efFects.
6 It is important to realize that even the pnctic:e of a legislator actively promoting
measures for his direct financial benefit is not necessarily corrupt; this was a common and
accepted state ofaffairs in Victorian England. See R.. A. Lewis, Ed"';" Cud",;,i au the
PU/U H,IIlt" MtIWIIIeIII1832-18S4 (Longman., Green, 1952), ch. 15, 'Reaction" which
Conclusion: The Future ofScience 419
affairs, or completely cynical, or capable of some sort of double
morality, he may maintain his personal integrity. But more com-
monly, a shadowy awareness that things are not quite as they must
be claimed to be, will force the individual agent to recognize the
possibility of his complicity in something culpable. Such recognition
tends to preserve and intensify the corrupted state of the activity:
fear of exposure comes to dominate the purposes of the agent, and
the group as a whole is held together by mutual blackmail. In such
a situation, the worst elements gain power over the better, and the
performance of the professed social functions is the least of the
considerations affecting individual and collective decisions.
It is easy to see that this sketched analysis applies to cases of
corruption in public life, where the private goals are the venal ones
of personal gain. On the other hand, it is by no means necessary that
every bureaucracy in which the defining functions have been dis-
placed is corrupt; if there is no significant public which had some
trust in it to begin with, there has been no betrayal. But again, it is
possible for officials of a voluntary or political organization to be-
come corrupted without desiring or achieving any personal benefit,
merely by finding it impossible to achieve the purposes of their
members and also finding it impossible to confess their failures.
And most tragically, it is possible for a man to discover in retro-
spect, after years of service to an organization, that he had all
along been corrupt. 7
Cases of corruption in technical projects can be quite straight-
forward: a public contract for a device is sought and procured, on
the basis of promised operating characteristics which the contractor
has neither the ability nor the intention of achieving, but where the
failure of the project will not affect his interests adversely. To pro-
tect his interests, the contractor may find it necessary to corrupt the
State agencies of control, so that they will merely pretend to
or
describes the successful campaign in Parliament to defend the privileges the Lcmdoa
water companies against the needs ofthe population. Even in the early twentieth century
Lincoln Steffens found English politicians who denied the existence of corruption in
England calmly desaibing practices which in America are certainly amsidered corrupt.
In trying to explain the differences between Europe and America he invoked the idea of
the 'old' and the 'new' civilizations, with 'corruption' as a natural historical process. Sec
Autobiography, Part IV, ch. VIII.
7 The type-case of unwitting corruption is Steffens's Captain Schmittberger, a
German immigrant who simply never knew that the policeman's job does fUJI include
protecting rackets and taking bribes, until the Lexow investigation exposed the truth
about the system, to himself. See Autollio"'.'''1, Part II, cbs. XII and XIII, pp. 266-8+
420 Conclusion: The Future ofScience
scrutinize his operations. 8 The situation is similar in practical projects:
the construction of a research empire may be publicly justified by its
function in the solution of urgent practical problems, while the
effective goals of the project are the provision of jobs, the securing of
an academic base, and the production of titles of publications. Even
in academic science, the production and publication of shoddy work
involves an element of corruption, since both author and referee are
participating in a deception, albeit an ambiguous one before a
largely anonymous public. Entrepreneurial science is by its very
nature corrupt in this sense, and an immature field in a hyper-
trophic state can scarcely avoid corruption. We notice, however,
that as we move away from the straightforward situations where
bank-notes are passed in return for favours, the subtlety and
ambiguity of corruption become more pronounced. Indeed, it is
possible for one person to denounce a project as corrupt, and another
indignandy and sincerely to deny it, the disagreement resting on
questions of whether there is an interested public, a trust, and a
betrayal.
A failure to come to terms with the new problems resulting from
the industrialization of science can bring about a very subtle but
none the less corrosive form of corruption within science as a whole.
For science, as a part ofacademic scholarship, has long claimed to be
discharging a public trust in the advancement and diffusion of
knowledge; and it has claimed a variety of privileges and immunities
for its members (not shared by other teachers or research workers)
on the basis of its ethic of autonomy and integrity. These claims are
different in character from those claimed by a learned profession,
and in some ways more extreme: a professional is expected to use
his judgement in solving the problem set by the welfare of the
client, while the scientist or scholar claims the freedom to choose the
problem itself. If it were possible to make a neat separation between
the sections of the scientific or academic community which are
devoted to scientific problems on the one hand, and technical and
practical problems on the other, then each section could develop its
distinct identity, with appropriate public justification of its position
and appropriate methods of social behaviour. But the industrializa-
tion of science brings the different sorts of problems into ever closer
connection, in institutions and in the work of individuals. There is
naturally a great temptation for the leaders of science to attempt to
• See Nieburg. /" tht N"mt ofSdmtt, for a discussion of this process.
Conclusion: The Future ofScience 42 1
gain the best of both worlds for as long as possible: to extol the
virtues of the free search for truth to one audience, and to promise
useful services to another. Except in those happy coincidences when
the different criteria ofvalue yield identical choices in a field, such a
situation is liable to produce corruption. The most easily identi-
fiable situation with such tendencies is what Americans call 'bootleg'
research, where resources obtained for one project are partly, at
least, diverted to another of more interest to the investigator. If this
is something that 'everyone does' in a particular community, the
equivalent of 'fiddling', then it is not corrupting to those involved. 9
On a large scale, however, as in big entrepreneurial science it can
have serious consequences.
A more subtle but more dangerous sort of corruption can occur
when the balance of real and professed final causes is tipped the
other way: when scientists claim the privileges appropriate to the
heirs of Helmholtz, while accumulating personal wealth and institu-
tional power through the regular contracting of mission-oriented
research. Again, the corrupting effects of such a situation may be
latent, until it is exposed and challenged. The natural response is
then to hide what can still be hidden and to explain away what
cannot. Up to now, such exposures, and their attendant crises, have
occurred only in connection with dirty science in American uni-
versities. Once the dangers of this situation were recognized, the
response of many leaders ofthe scientific community was admirabl~.
They disengaged from the State in military scientific establishment,
doubtless at considerable personal cost, when the Vietnam war
became politically and morally indefensible, and even before militant
students forced the issue at leading universities. However, such a
move, although welcome and heartening in itself, does not resolve
the underlying dilemma of the external relations of industrialized
science, with its tendency to corruption. to
9 See D. s. Greenberg, '''Boodeggingn: it Holds a Firm Place in Conduct ofResearch',
Snmet, 153 (19 September 1966), 848-9. With cbaracteristically American sophistication,
some large industrial research laboratories become worried if their scientists work only
on the projects formally agreed upon.
10 Some scientific communities maintain their independence and integrity by
utonishingly direct means. In Japan, physicists who associate themselves with the Japan
Defence Agency are ostracized by the Japan Physical Society, not being permitted to
Pl\*Ilt papers at its conferences. Military personnel sent to graduate schools by the
armed forces are failed on their exams, either on entrance or on completion of their
coone. In these circumstances, it is not surprising that 'Japanese defence oflicials have
also privately admitted to American colleagues that they have difficulty getting scientists
IS-S.te.S.P.
422 Con&lwion: The Future ofScience
The State, and industry, need an expertise more sophisticated and
prestigious than can be provided by narrowly technical institu-
tions and personnel; it must come from the world of science based
on academic institutions. I I And if it is to continue to receive support
on the present large scale, science must provide this service. To do
so effectively, in the Anglo-American institutional structure, re-
quires science to maintain a plausible semblance of its autonomy and
integrity; yet the autonomy of science cannot be more than a
semblance, especially as its accountability to its paymasters becomes
increasingly close and obvious. The world of science will then need
to half-believe itself to be still academic, free and autonomous, while
half-knowing itself to be industrialized, dependent, and responsible
to the State and industry. The traditional professed functions of
science are internal and under its control: a means to the ultimate
goal of the advancement of knowledge, as conceived and guided by
itself. But the actual goals of the work are increasingly subordinate
to functions that are externally defined: a means to the fostering of
civil and military industry, through the provision of particularly
applicable results and trained experts, on demand. When and
whether this divergence will become a part of the self-consciousness
of science, and by whom it could then be considered as a betrayal of
a public trust, depends entirely on the complex circumstances of
history and society. In this connection, one may revise Lord Acton's
aphorism, 'all power corrupts, and absolute power corrupts abso-
lutely', and substitute, 'responsibility tends to corrupt, and respon-
sibility without power corrupts absolutely'.12
Critical Science: Politics and Philosophy
We can now permit ourselves some final speculations on possible
ttends in the future of the natural sciences. The process of industri-
to pcrfonn defence research'. See P. M. Boffey, 'Japan (I): On the Threshold of Big
Science ?', Scim&t, 16, (1970), 31-5.
II 'Precisely the indiscipline ofrelatively free intel1ectual activity attracts the powerful
u guaranteeing the relative disinterestedness and novelty they hope to find in the ideas
of intel1ectual counse11on. It is one of the ironies of our time that so many intellectuals
strive to identify with the penpectives of kings, whilst their rulers value them for their
activity as philosophers.' See N. Birnbaum, 'On the Idea of a Political Avant-garde in
Contemporary Politics: the Inte11ectuals and Technical Intelligentsia', Praxis (Zagreb,
1969), Nos. 1-2, p. 343·
12 This aspect of corruption is discussed in my papers on 'Power, Responsibility,
Answerability'. The context there was the problem of participation in university
government; but it could be interesting to relate it to the position ofscientists responsible
to the State described by the Amaican c1ich6, 'on tap but Dot on top'.
Conclusion: The Future ofScience 423
alization is irreversible; and the innocence of academic science
cannot be regained. The resolution of the social problems of science
created by its industrialization will depend very strongly on the
particular circumstances and traditions of each field in each nation.
Where morale and effective leadership can be maintained under the
new conditions, we may see entire fields adjusting successfully to
them, and producing work which is both worthwhile as science and
useful as a conttibution to technology. Recruits to this sort ofscience
will see it as a career only marginally different from any other open
to them; and it is not impossible for men of ability and integrity to
rise to leadership in such an environment. This thoroughly in-
dusttialized science will necessarily become the major part of the
scientific enterprise, sharing resources with a few high-prestige
fields of 'undirected' research, and allowing some crumbs for the
remnants of small-scale individual research. A frank recognition of
this situation will help in the solution ofthe problems ofdecision and
conttol. Since the criteria of assessment of quality will be heavily
biased towards possible technical functions of results, they will
thereby be more easily applied, and less subject to abuse, than those
which are based on the imponderable 'internal' components of
value.
Thus, provided that the crises in recruitment and morale do not
lead to the degeneration and corruption of whole fields, we can
expect emergence of a stable, thoroughly industrialized natural
science, responsible to society at large through its conttibution to
the solution of the technical problems set by industry and the State.
Scientists, and their leaders and institutions, will be 'tame': accept-
ing their dependence and their responsibilities, they will be unlikely
to engage in, or encourage, public criticisms of the policies of
those institutions that support their research and employ their
graduates. Such a policy of prudence is not necessarily corruption;
whether it becomes so will depend on many subde factors in the
self-consciousness of this new sort of science, and the claims made
to its audiences. But not all the members of any group are easily
tamed, and the emergence of a 'critical science', as a self-conscious
and coherent force, is one of the most significant and hopeful
developments of the present period.
There have always been natural scientists concerned with the
sufferings of humanity; but with very few exceptions they have
faced the alternatives of doing irrelevant academic research to gain
424 Ctm&lusion: The Future ofScience
the leisure and freedom for their social campaigns, or doing applied
research which could benefit humanity only if it first produced
profits for their industrial employer. The results of pharmaceutical
research must pass through the cash nexus of that industry before
being applied, and that process may be an unsavoury one. Only in
the fields related to 'social medicine' could genuine scientific research
make a direct conttibution to the solution of practical problems, of
protecting the health and welfare ofan otherwise defenceless public.
Now, however, the threats to human welfare and survival made by
the runaway technology of the present provide opportunities for
such beneficial research in a wide range of fields; and the problems
there are as difficult and challenging as any in academic science.
These new problems do more than provide opportunities for scien-
tific research with humanitarian functions. For the response to this
peril is rapidly creating a new sort of science: critical science. In-
stead of isolated individuals sacrificing their leisure and interrupting
their regular research for engagement in practical problems, we
now see the emergence of scientific schools of a new sort. In them,
collaborative research of the highest quality is done, as part of
practical projects involving the discovery, analysis and criticism of
the different sorts of damage inflicted on man and nature by run-
away technology, followed by their public exposure and campaigns
for their abolition. The honour of creating the first school of 'critical
science' belongs to Professor Barry Commoner and his colleagues at
Washington University, St. Louis, together with the Committee for
Environmental Information which publishes Environment. 13 For
some years a Society for Social Responsibility in Science, based in
America but with members and branches overseas, was the main
voice of conscience in science; recently a British Society for Social
Responsibility in Science, with a rather broader base among the
leaders of the national scientific community, has been fonned. As
such societies gain sttength and influence, the success of the St.
Louis school of critical science should soon be emulated elsewhere.
The problem-situations which critical science investigates are not
the result of deliberate attempts to poison the environment. But
they result from practices whose correction will involve inconveni-
ence and money cost; and the interests involved may be those of
powerful groups of firms, or agencies of the State itself. The work
I J The first statement of 'aitical science' as distinct from ecological concern is
B. Commoner, Stimtt ,,1Ul Swvi.l (Gollanc:z, London, 1966).
Conclusion: The Future ofScience 42S
of enquiry is largely futile unless it is followed up by exposure and
campaigning; and hence critical science is inevitably and essentially
political. 14 Its style of politics is not that of the modem mass move-
ments or even that of 'pressure groups' representing a particular
constituency with a distinct set of interests; it is more like the
politics of the Enlightenment, where a small minority uses reason,
argument, and a mixture ofpolitical tactics to arouse a public concern
on matters of human welfare. IS The opponents of critical science
will usually be bureaucratic institutions which try to remain faceless,
pushing their tame experts, and hired advocates and image-
projectors, into the line of battle; although occasionally a very
distinguished man is exposed as more irresponsible than he would
care to admit. 16
In the struggles for the exposure and correction of practices
damaging the environment the role of the State is ambiguous. On
the one hand, every modem government is committed in principle
to the protection of the health of its people and the conservation of
its natural resources. But many ofthe agencies committing the worst
outrages are State institutions, especially the military; and in any
event the powerful interests which derive profit or convenience from
polluting and degrading the environment have more political and
economic power than a scattering of 'conservationists'. It some-
times occurs that two State agencies will be on opposite sides of an
environmental sttuggle; but the natural tendency ofregulatory agen-
cies to come under the control of those they are supposed to regulate
can make such a struggle a one-sided affair.
The presence of an effective critical science is naturally an
embarrassment to the leadership of the responsible, industrialized,
14 The most comprehensive analysis of 'aitical science' yet published is MuNichoJson.
Tile EmJirtnmle1ltlll Revolution (Hodder & Stoughton, London, 1970). He is mainly
concerned with 'conservation', but his healthy approach to modem bureauaatic politics
is developed in his earlier book, Tile System (Hodder & Stoughton, London, 1967;
McGraw-Hill. 1969).
15 Support for this new style of politics bu come from the Duke of Edinburgh.
Speaking on the B.B.e. programme '34 Hours', on 17 February 1970, he discuaed the
proposition that 'tough action against the poisoners and wrecken' was essential for the
promotion ofconservation and the abatement of pollution, and said that people must 'be
ruder and more direct to the people in political authority'.
16 On 22 March 1966, for example, the President of General Motors appeared before
a Senate hearing to apologize to Mr. Ralph Nader, following revelations that General
Motors' lawyers had hired an investigator to unearth details about Nader's private life.
Nader's analysis of the defects of the 'Conair' car was aJSting General Moton a lot of
money.
426 Con&lusion: The Future ofS,imce
tame scientific establishment. l ? Their natural (and sincere) reaction
is to accuse the critics of being negative and irresponsible; and their
defensive slogan is along the lines of 'technology creates problems,
which technology can solve'. This is not strictly ttue in all cases,
since nothing will solve the problems of the children already killed
or deformed by radioactive fallout or by the drug Thalidomide. ls
Moreover, this claim canies the implication that 'technology' is an
autonomous and self-correcting process. This is patent nonsense.
We have already seen that a new device is produced and diffused
only if it performs certain functions whereby human purposes can
be served; and if the intended beneficiaries do not appreciate its use,
or if those injured by its working can stop it, the device will be still-
born. The distortions of technological development arise when the
only effective 'purposes' in the situation are those of the people who
believe themselves to derive pure benefit from the innovation. On
the self-correcting tendency of technology, one might argue that no
large and responsible institution would continue harmful practices
once they had been recognized; but this generalization is analogous
to the traditional denial of the cruelty of slavery, along the lines that
no sensible man would maltreat such valuable pieces of property.
And the history of the struggles for public health and against pol-
lution, from their inception to the present, shows that the guilty
institutions and groups of people will usually fight by every means
available to prevent their immediate interests being sacrificed to
some impalpable public benefit. 19 If the campaigns waged by
mtical science come to touch on some issue centtal to the conven-
ience of the State or other very powerful institutions, we may
experience a polarization of the community of natural science,
along the same lines as has already occurred on the Vietnam issue in
some of the human sciences in America. In such a situation, it will
17 Thus N"""e ridiculed the UNESCO amference on the biosphere of September
1968t comparing it to an earlier amference on 'communications satellites and under-
deft10ped countriest. See 'Bandwapn for UNESCO't N.,,,,e, 219 (7 September 1968)t
p. 999. There was DO report on the amference itself:
II Up to the time ofwritingt N"""e maintains a magnificent complacenCYt such u one
bardIy hopes to see in England in i1B epoch of imperial decline. Thus, aiticizing Dr.
Fruer Darling'. fint Reith Lecture. an editorial ~ 'And in spite of quite proper
concern for the need to make ouly decent use of new developments, it is hard to find
contemporary i llustntions of where teclmology has gone astray.' 'No Peace for the
Wicked t t N"""e, 234 (1969), 63 1 •
19 The 'IIDitary movemeni in nineteeDth-century England was involved in such
IttUp through i1B career. For a ample, see It A. Lewis, U",;n Cluulwi tIIUl ,lie
p"jlie HMlII MIIfJaIeIII 1832-S4 (Longmans, Green, 1952).
Conclusion: The Future ofScience 427
not be possible for a leader of science to be both honest and tame;
and if the establishment of science chooses to serve its paymasters
rather than ttuth, it will be recognizably corrupt.
Such extreme situations may be a long time in developing, if for
nothing else than that critical science is still in its infancy. As it
develops, it will be at risk of encountering many pitfalls, partly
those characteristic of immature sciences applied to practical prob-
lems, and partIy those of radical and reforming political move-
ments. Perhaps the most obvious will be an accretion of cranks and
congenital rebels, whose reforming zeal is not matched by their
scientific skill. But there are others, arising from the contradictory
relations between critical science and the relevant established in-
stitutions of society. As ttue intellectuals rather than a technical
intelligentsia, individual members may find some 'sinecures within
the interstices of bureaucratized intellectual systems'; 20 but there
will need to be some institutions providing a home for the nucleus
of each school, and external sources of funds for research. Hence,
especially as critical science grows in size and influence and society
becomes more sophisticated about the problems of runaway tech-
nology, some accommodation between the critics and the criticized
will inevitably develop. We can even expect to see critical research
being supported, critical slogans being echoed, and leaders of
critical science being rewarded, by institutions whose basic des-
tructive policies still are unchanged. 21 Such phenomena have
already occurred in America, in the politics of race; and on this
issue, where the interests concerned are mainly major institutions
which can hire talented and enlightened experts at will, it is even
more likely. The movement of critical science would then face the
pitfalls of corruption as soon as, or even before, it had skirted those
of impotence. 22 But this is only a natural process, characteristic of
20 ' ••• the intellectuals' distance from certain kinds of material activity, their occupa-
tional repugnance for certain forms of boUl'pOis organizations, their attachment to
abstract versions of bourgeois tradition nther than to the lUb-stratum of bourgeois
activity, their familiar quest for sinecures within the intentices of bureaucratized
intellectual systems, combine to endow them with what was once an anticapitalist and is
now an anti-bureaucratic ethos.' See N. Birnbaum, Ope cit., p. 24+
31 First prize in the 'enlightenment' stakes bu been won by the Mcmsanto C1emical
Company. The Scientific Division of the Committee for Environmental Information
(which publishes EmJirtnmle1lt) included among its members for 1969 Mr. F. D. Wharton,
Jr., St. Louis Development Manager, Life Sciences in the New Enterprise Division,
Monsanto Company.
23 On the corruption of good causes, the classic is G. B. Shaw, M"jor BM••• The
climu of the play comes when Mrs. Baines, the Salvation Army (Ammissioocr who
428 Conclusion: The Future ofScience
all radical movements. It is easy to maintain one's integrity when
one's words and actions are ineffective; but a long period of this can
produce a sectarian or a crank. If one begins to achieve power, and
one's policies affect the interests of many others, one must decide
where one's responsibility lies. If it is to the ideal alone, then one is
set on a course towards tyranny, until overthrown by the host of
enemies one has raised Up.23 And if one accepts responsibility for
the maintenance of a general welfare, including that of one's
opponents, one is on the path to corruption and impotence. This
may seem a gloomy prognosis; but a society which does not present
such hazards to radical movements of every sort is not likely to
retain its stability.
We can expect, then, that the future political history of critical
science will be as complex and perhaps as tortured as that of any
successful radical and reforming movement. But if it does
survive the pitfalls of maturation, and so contributes to the
survival of our species, it can also make a very important con-
ttibution to the development of science itself. For if the style
of critical science, imposed by the very nature of its problems,
becomes incorporated into a coherent philosophy of science, it
will provide the basis for a transformation ofscientific inquiry as
nIDI the shelter in West Ham, tbanb God for the douation of {.s,ooo by Mr. Bodger,
the distiller whose whiskey is the curse of the poor in their care. The Cockney Bill
Walker, whose donation of a guinea in repentance for striking two women bad just
previously been indignandy rejected, utters the significant 'Wot prawce selvytion nab?'
(penguin Boob, 1960; p. 106.) Sec also Shaw's Preface to the play.
:&J A cautionary tale that should be read by aU who are embarking on political activism
baed on 'aitical science' is the play by Ibsen, Tile £""'" oftile PtOple. Superficially, it is
about an honest doctor who is hated by the corrupt forces of his town for his determina-
tion to expose the scandal of polluted waters being used in the town's profitable baths.
But on closer reading, it can be seen that Dr. Stoebnann'. misfortunes were also due to
his own D&i~ and egoism. It is significant that in his own version of the play (Viking
Praa, New York, 1951), Arthur Miller strengthened the 'progressive' message by trans-
ferring the paaap where the public meeting declares Dr. Stockmann to be 'an enemy of
the people'. In his version it comes at the very beginnins of the meeting, before he has
spoken; while in the original it comes after the Doctor'. harangue, amcluding with, 'Let
the whole country perish, let all these people be exterminated.' Mer studying the play
with a class at Harvard, where this modification was discovered, it struck me that a
worthwhile sequel could be written, entided 'The People'. Friend', in which the
entrenehed forces, ifouly a bit lea stupid and venal than in the original, could corrupt the
good Doctor without difticulty. Although Spa resorts are DO longer an important focus of
pollution, it is possible for an 'Enemy of the People' situation to be repeated in any of the
seaside toWDS which dump their nw sewage into proDmity to bathers. See J. A. Wakefield,
'Qean or Dirty Beaches-Which do you Prefer?' Y"" ErrvirtnIIMIII, No. 1 ~mtel'
1969), pp. 29-3 1•
Conclusion: The Future ofScience 429
deep as that which occurred in early modem Europe. The
problems, the methods, and the objects of inquiry of a matured
and coherent critical science will be very different from those of
academic science or technology as they have developed up to now;
and together they can provide a practical foundation for a new
conception of humanity in its relations with itself and the rest
of nature.
The work of inquiry in critical science involves an awareness of
craft skills at all levels, and tile conscious effort of mastering new
skills. The data itself is obtained in a great variety of ways, from the
laboratory, from the field, and from searching through a varied
literature, not all of it in the public domain. Much of it lacks sound-
ness, and all of it requires sophisticated and imaginative treatment
before it can function as information. Indeed, since the problem-
situations are presented in the environment, and much of the
crucial data must be produced under conttolled conditions in the
laboratory, work in critical science may overcome the dichotomy
between field-work and lab-work which has developed in science,
even in the biological fields, over the past century. In the later phases
of investigations of problems, the same challenges of variety and
novelty will always be present. The establishment of the sttength
and fit of each particular piece of evidence is a problem in itself;
and the objects of inquiry (including the measures of various effects
and processes, as well as conventional standards of acceptability in
practice) are so patently artificial, that one is in little danger of
being encased in them as a world of common sense. The establish-
ment of criteria of adequacy for solved problems is possible, for the
work will frequently be an extension and combination of established
fields for new problems, and so critical science can escape the worst
perils of immaturity. Also, any critical publication is bound to be
scrutinized severely by experts on the other side, so high standards of
adequacy are required because of the political context of the work.
Indeed, a completely solved problem in critical science is more
demanding than in either pure science or technology. In the former,
it is usually sufficient to obtain a conclusion about those properties
of the artificial objects of inquiry which can be derived from data
obtained in the controlled conditions of experiment; in the latter it
is sufficient for an artificial device to perform its functions without
undue disturbance by its natural environment; while here the com-
plex webs of causation between and within the artificial and natural
430 Ctm&lusitm; The Future ofScience
systems must be understood sufficiently so that their harmony can
be maintained.
The social aspects of inquiry in critical science are also conducive
to the maintenance of its health and vitality, at least until such times
as the response to its challenge becomes over-sophisticated. The
ultimate purpose which governs the work is the protection of the
welfare of humanity as a part of nature; and this is neither remote,
nor vulgar. Critical science cannot be a permanent home for career-
ists and enttepreneurs of the ordinary sort; although it may well use
the services of bright young men intending eventually to serve as
enlightened experts. Those who want safe, routine work for the
achievement of eminence by accumulation will not find its atmos-
phere congenial; for its inquiries are set by a succession of problem-
situations, each presenting new challenges and difficulties. Hence
although critical science will doubdess experience its periods of
turbulence, political and scientific, it is well protected from stagna-
tion and from the sort ofcreeping corruption that can easily come to
aftliet industtialized science.
Finally, the objects of inquiry of critical science will inevitably
become different from those oftraditional pure science or technology,
for here the relation of the scientist to the external world is so
fundamentally different. In traditional pure mathematical-experi-
mental natural science, the external world is a passive object to be
analysed, and only the more simple and abstract properties of the
things and events are capable of study. In technology, the reactions
of the unconttolled real world on a constructed device must be
taken into account, but only as perturbations of an ideal system; the
task is to manipulate it or to shield the device from its effects. But
when the problem is to achieve a hannonious interaction between
man and nature, the real world must be tteated with respect: both
as a complex and subtle system in its own right, and as a heritage of
which we are temporary stewards for future generations. Hence,
even though studies of our interaction with the environment will
necessarily use all the intellectually constructed apparatus of dis-
ciplined inquiry, their status and their content will inevitably be
modified. They will be more easily recognized as imperfect tools,
with which we attempt to live in harmony with the real world around
us; and although this attitude may seem to conduce to scepticism,
it will be the healthy one which recognizes that genuine knowledge
arises from lengthy social experience, and that such knowledge
Conclusion: The Future ofScience 43 1
depends for its existence on the continued survival ofour civilization.
The objects of inquiry themselves will include final causes among
their essential attributes, not merely the limited functions appro-
priate to technology, but also the judgements of fitness and success
already developed in classical biology and ecology. All this is work
for the future; but if it is successful, the opposition between
scientific knowledge and human concerns, characteristic of the
sciences derived from the dehumanized natural philosophy of the
seventeenth century, will be overcome.
With a new conception of the practice of science, there will come
a new conception of the history of science and of the meaning of the
scientific endeavour. It is possible, and it has been natural, to
reconstruct the history of scientific inquiry as a success-story
leading up to the ttiumph of the matured academic mathematical-
experimental natural science of the later nineteenth century. One
can identify the historical moments, and the great men associated
with them, when the very conception of the inqUiry itself was
significantly advanced. Thus, the origins of our sort of science are
rightly located. in the earlier Greek civilization, when attempts were
made to account for the world of sense experience without invoking
personified divine agents. The heritage ofthe so-called 'pre-Socratic'
philosophers was further developed by Aristotle, who not merely
conducted disciplined inquiries over almost all areas of human ex-
perience, but also showed that such disciplined inquiry has conceptual
and methodological problems that can and should be investigated. In
a parallel tradition, the idea of mathematics as a body of proved
results about conceptual objects developed to full maturity in the
achievements of Euclid and Archimedes. The next great advance
(according to this interpretation) came many centuries later, 'when
the pioneers of the 'mechanical philosophy' of the seventeenth
century achieved a powerful synthesis of experience and reason.
GaliIeo's slogan, 'sense experience and necessary demonstration',
stood for his appreciation of the need for closely controlled experi-
ence, which could serve as evidence of the appropriate strength, in
an argument cast in mathematical language. All that was required to
complete the body of methods of classic academic science was the
development of institutions for organized, co-operative research;
and this came by natural evolution through the nineteenth century.
The dominant world-view of matured academic science was atom-
istic in several important senses. The real world underlying our
432 Ctm&lusion: The Future ofScience
sense experience was assumed to be devoid of the characteristically
human attributes of intellectual and spiritual reality, and of value;
final causes were excluded from scientific explanation; and all
efficient causes were to be reduced to the material cause of insensibly
small brute matter in motion. Correspondingly, knowledge itself
was atomic: the achievement of individual facts about the external
world, isolated from any philosophical and social context, was
considered possible and valuable. This approach to natural know-
ledge achieved magnificent successes in many fields, and was also
appropriate for the development of successful large-scale research.
It was natural to suppose that this particular style of scientific
enquiry could be successfully extended to all disciplines, and that it
was internally stable. But both these optimistic assumptions proved
incorrect. Ineffective and immature fields can be hindered rather
than helped by a mechanical imitation of those whose objects and
appropriate methods are very different; and the pretence of this sort
of maturity only adds to the hazards of applying such fields to
practical problems. On the other hand, academic natural science has
been transformed by its very successes into industrialized science;
and the unexpected problems and abuses of this new sort of science,
ranging from shoddy science to runaway technology, present
threats to the survival of science and of our whole civilization.
With the new perspective gained from our recent experience, we
can look again at the long history of the human endeavour of under-
standing and controlling the external world. We can now see a
positive significance in events and tendencies that have hitherto been
considered as unfortunate aberrations. The dominant traditions in
academic science have developed out of conflict with other styles of
scientific work; and it can be distinguished from them by its objects
of inquiry, its methods of work, and its social context. For 'our'
science, the real world is devoid of human and spiritual qualities;
and the scientist studies the smaller aggregations of matter, con-
sidering the most simple and mathematical properties that suffice
for the successful investigation of problems. Its approach to its
materials is appropriately depersonalized; any 'deeper' meaning
that might be thought to inhere in its results is rigidly segregated
from his reporting, and is left to amateur speculations. As a social
activity, this science is necessarily elitist, presupposing a lengthy
course of training and indocttination for which only a minority have
an appropriate cultural background.
Conclusion: The Future 0/ Science 433
To the extent that the ttaditional history ofscience has considered
these aspects of scientific inquiry, it has been embarrassed by the
presence of traditions and tendencies that achieved success in 'our'
terms in spite of radical differences in one or more respects from the
recently dominant academic style. The roots of astronomy in
astrology, and of chemistry in alchemy, are cases in point. Some of
the immortal ancestor-figures of the modem discipline are revealed,
on unbiassed inspection, to have seen their work as conttibuting to
what is now regarded as pseudo-science: Ptolemy and Tycho for
astronomy, and Paracelsus and Glauber for chemistry. Indeed,
when we look more closely at the period of the later sixteenth
century, when the arts and sciences were developing quite rapidly
he/ore the incursion of the 'new philosophy' of dead matter, we find
the very greatest scientists participating in the world-view of an
animated nature: Gilbert investigating magnetism in the attempt to
prove that the earth is the embodiment of the anima mundi, Kepler
searching (with all rigour) for the divine harmonies of the celestial
realm, and Harvey using 'spirit' and the macrocosm-microcosm
analogy to guide his anatomical and physiological researches. Z4
It would be very misleading to imagine a simple succession oftwo
sorts of science, each unified and coherent in itself, first that of the
'animated' world and then (since the seventeenth century) that of
the 'dehumanized and disenchanted' world. History is more complex,
and more interesting, than that; and within the 'old' conception of
science there were many different tendencies in the interpretation
of its appropriate objects, methods and social functions. I have
previously referred to a 'romantic' philosophy of nature providing
the vehicle for a politically radical folk-science that challenges the
academic science of its time. In this tradition, the study of nature is
explicitly seen as a social and also spiritual aet; one dialogues rather
than analyses; and there is no protective cover of belief in the
'neutrality' or 'objectivity' ofthe style adopted. Such a philosophy of
nature will become articulated and advanced, as part of a general
radical reaction against a formal, dry or bureaucratic style pervading
social or cultural life. Looking back into history, we can find a
similarity of doctrine or style, and sometimes a linking ttadition, as
24 Gilbert makes his programme plain; see tie M"pete Book V, ch. 12: 'The magnetic
force is animate, or imitates a soul; in many respects it surpuses the human soul while
that is united to an organic body! (b'. P. F. MotteIay; Dover, New York, 1958, p. 308.)
Kepler is well-known; and for Harvey, see W. Pagel, WillUJ", H1lrW1'1 BioloF'" Itle.,
(Buel and New York, 196'7).
434 Ctm&lusion: The Future ofScience
far back as the Taoists of ancient China, through St. Francis of
Assisi, to Paracelsus, William Blake, and Herbert Marcuse. 25
Not every one of these figures would claim to be a natural
scientist of any description; but as philosophers, poets or prophets,
they must be recognized as participating in and shaping a ttadition
of a certain perception of nature and its relation to man. Granted all
the variety of their messages and styles, certain themes recur. One
is the 'romantic' striving for immediacy, of contact with the living
things themselves rather than with book-learned descriptions.
Another is 'philanthropy'; the quest is not for a private realization,
but for the benefit of all men and nature. And, related to these is a
radical criticism ofexisting institutions, their rules and their person-
nel. Looked at from the outside, each upward thrust of the romantic
philosophy of nature is doomed to failure. Mankind. will not be
transfigured overnight; and the romantic style has its own destruc-
tive conttadietions. Whereas the 'classic' style degenerates gradually
into an ossified form and a sterile content, the 'romantic' style goes
off much more quickly, through chaos of form and comfption of
content. But this study of ours has shown that even in disciplined
scientific inquiry, the categories of 'success' and 'failure' are neither
so absolutely opposed, nor so assuredly assignable in particular cases,
as the ttaditional ideology of science assumed. And the failure to
achieve Utopian dreams, in science as well as in social reform, is
not at all the same thing as futility.
The dreams of the romantic, philanthropic, radical philosopher-
prophets cannot move towards realization by the accumulation of
facts or of battalions. Rather, they exist through a discontinuous,
perhaps erratic, series of crises and responses. Sometimes they have
the good fortune of producing a creative tension in a man brave
enough to attempt the synthesis of a prophet's vision with a world
as On Taoism, see J. Needham, Seienee .rul CiviliSlltitm in Chi"", 2, 88-132. In his
magisterial fashion, Needham provides more materials on the analogous movements in
early modern Europe than is available in any general history ofscience. For a discussion
of the limitations of his view, see note 35 on p. 394 above. See Lynn White, Jr., 'The
Historical Roots ofourEcological Crisis', MII&hine ez/)eo (M.I.T. Press, 1968),Chapter 5,
for St. Fnncis.
Fnncis A. Yates, in 'The Hermetic Tndition in Renaissance Science' in ~t, StilfU'e
all HilttW:! ill tile ReuiSSllll&e, ed. C. S. Singleton Oohn Hopkins Press, Baltimore, 1968),
discusses the 'Rosicrucian' style ofscience in considerable depth. The theme.of 'philan_
thropy' is most clearly developed in the German alchemical philosophers in the Paracel-
Iian tradition; and their influence on Francis Bacon is clear.
On Marcuse, see his One-Dimmsitnull M." (Bcac:on Press, Boston, 1964).
Conclusion: The Future ofS&imte 435
managed by priests. He too will fail, almost certainly; some problems
are insoluble. But his message, perhaps in a particular science or
walk of life, perhaps of a generalized wisdom, will speak to men in
later ages, coming alive whenever it has insights to offer. In this
present period, we may find Francis Bacon speaking to us more than
Descartes the metaphysician-geometer or Galileo the engineer-
cosmologist. As deeply as any of his pietistic, alchemical forerun-
ners, he felt the love of God's creation, the pity for the sufferings of
man, and the striving for innocence, humility, and charity; and he
recognized vanity as the deadliest of sins. 26 To this last he ascribed
the evil state of the arts and sciences:
For we copy the sin of our first parents while we suffer for it. They
wished to be like God, but their posterity wish to be even greater. For we
create worlds, we direct and domineer over nature, we will have it that all
things are as in our folly we think they should be, not as seems fittest to the
Divine wisdom, or as they are found to be in fact. 27
The punishment for all this, as Bacon saw it, was ignorance and
impotence. It might seem that the problem is different now, for we
have so much scientific knowledge and merely face the task of
applying it for good rather than evil. But Bacon assumed his readers
to believe themselves in possession of great knowledge; and much of
his writing was devoted to disabusing them of this illusion. Perhaps
the daily reports of 'insufficient knowledge' of the effects of this or
that aspect of the rape of the earth, and our sense of insufficient
understanding of what our social and spiritual crises are all about,
indicate that in spite of the magnificent edifice of genuine scientific
knowledge bequeathed to us, we are only at the beginning oflearning
the things, and the ways, necessary for the human life.
Bacon was a shrewd man, fully sensitive to the weaknesses of the
human intellect and spirit. He was aware of the superficiality of
ordinary thought and discourse, at whatever educational level; and
he also distrusted the extraordinary enthusiast, in religion or politics,
for the damage he could cause. His life's endeavour was to overcome
26 For a detailed interpretation ofBacon'. programme for scipnce in terms of. vision
of moral and spiritual reform, see J. R. Ravetz, 'Francis Bacon and the Reform of
Philosophy', in Stimee, Mediane MIll SDtin;yi" tileReuiSllMee (Walter Pagel FestICbrift),
cd. A. Debus (University.ofChicago Press and Oldboume Press, London, 1972). This is
an elaboration of certain themes in Benjamin FarriDgton's, Tile PJaloSlllIl;y of p,,,.;,
Bilton, and I am indebted to him for my tint iDsights into this aspect of Bacon.
27 See Tile NlltWtU "rul Ezpm1llllll1ll HUttW,ftW tile F~ qfPIIiIoIfJ,II, (WtWa.
vol v; traD8Iation p. I~).
436 Conclusion: The Future ofScience
this conttadietion somehow, and to bring about a ttue and effective
reformation in the arts and sciences of nature. For him, this was a
holy work, a work of practical charity inseparable from spiritual
redemption. 28 His audience was inevitably among the literate; and so
he ttied, by scattering hints and half-concealed invitations, to call
together his brothers, who would gendy and silendy show by their
example that a good and pure way into Nature is also the practically
effective way. Of course he failed, in his philosophical reform as in
his political career. There was no English audience for his particular
message during his lifetime, and at his death he was alone and
neglected.
Shortly after his death, however, there was a stining; and Bacon's
message of 'phi1anthropic' science began a career of its own. For a
while, his followers knew what he was about; but with the passage
of decades and disillusion, this was forgotten, and only the vulgar
fact-finding Bacon survived. Yet when we now come back to read
Bacon, perplexed and worned as we are by the sudden ttansfor-
mation that science has wrought upon itself as well as upon the
world, we can find relevance in passages like the following:
Lastly, I would address one general admonition to all; that they con-
sider what are the true ends of knowledge, and that they seek it not either
for pleasure of mind, or for contention, or for superiority to others, or for
profit, or fame, or power, or any of these inferior things; but for the
benefit and use of life; and that they perfect and govern it in charity. For
it was from lust of power that the angels fell, from lust of knowledge that
men fell; but of charity there can be no excess, neither did angel or man
ever come in danger by it. 30
al See the MellillllUmes SIIerM (WtWis, voL vii; translation pp. 243-4). Bacon contrasts
the miracIa ofpunishment wrought by the prophets ofthe Old Testament, with those of
Jesus: 'Jesus was the Lamb of God, without wrath or judgment. All his miracIa were for
the benefit of the human body, his doctrine for the benefit of the human soul' After a
list of instances, Bacon comments, 'There was no miracle of judgment, but all ofmercy,
and all upon the human body.' Later, in the essay 'Of Hypocrites', he comments, 'The
way to convict • hypocrite therefore is to send him from the worb of sacrifice, to the
works of mercy. Whence the text: P"e relititm MUlllllllefiJetl1leftWe Coil "rullM F"/Mr
;s I.S, 10 tUIIM tWpluuu tllUllI1itlo",s ;n 11Ieir "J!Iie'itm •••' (p. 249).
at On the influence of Bacon, see Qwles Webster, S.",.l H.,liIJ "rul'M AtlWlnte-
,.", ofu.";,,, (Cambridge University Press, 1970).
JO Bacon, Till GrMlI",,1IW1JIitnJ, Pre/M' (Wora, vol iv; tnnsJation pp. 20-21).
INDEX OF NAMES
Here are listed definitions and extended discussions of basic concepts developed in this
work. Compiled with the assistance of Joseph Ravetz.
SCIENTD'IC KNOWLEDGE
AND ITS SOCIAL PROBLEMS
Jerome R. Ravetz
With a new Introduction by the author
Science Is continually confronted by new and d1fDcult social and ethical
problems. Some of these problems have arisen from the transformation of
the academic sclence of the prewar period into the Industrialized science of
the present. Traditional theories of science are now widely recognized as
obsolete. In ScI.entfIIc Knowledge and Its SocIal Problems (or1g1nally published
in 1971), Jerome R. Ravetz ana1yzes the work of science as the creation and
investigation of problems. He demonstrates the role of choice and value
judgment, and the inevitabWty of error, in sclentlflc research. Ravetz·s new
introductory essay Is a masterful statement of how our understanding of
sclence has evolved over the last two decades.
-SctentYfc Knowledge and Its SocIal Problems Is both monumental and remark-
able ... a major contribution to the understanding of sclence.- -Jonathan
Rosenhead, New Sctenttst and Sc1ence Journal
-mhoughtful, incisive, scholarly, lucid, humane and sane. Everyone-but
everyone-within or concerned about science should read It slowly and
carefully.- -John ZIman, Nature
-nus Is a penetrating and novel account of the sclentlflc activity. - - Ttmes
Uterary Supplement
-Scten.tf/lc Knowledge and Its Soc1.al Problems, Jerome R. Ravetz's major
contribution to the history and philosophy of sclence, should be of interest to
anyone concerned with change in education.- - Robert McClintock. Teachers
College Record
-Scten.tf/lc Knowledge and Its SocIal Problems deserves to stand for some time
to come as one of the landmarks of the new cI1t1cal spll1t which Is transfonnJng
science from within in ways whose political and cultural bnportance cannot
be exaggerated.- -Theodore Roszak. The Nation
About the AIIthor
ISBN 1-56000-851-2
90000>